When he cross-examined the Merlin Operation manager Bob S at Jeffrey Sterling’s trial, defense attorney Barry Pollock asked whether Bob S  thought he was doing 70% of the thinking on the operation. When Bob S denied that, Pollock reminded Bob S of his February 28, 2006 FBI testimony, where he had said he was doing 70% of the thinking to Sterling’s 30%. “This was shortly after publication of book that revealed the whole operation,” Bob S explained his earlier comment. “I was being ungenerous.”

Similarly, when he cross-examined Merlin himself, defense attorney Edward MacMahon asked whether he had told the FBI in March 2006 that Sterling (whom elsewhere Merlin called “lazy” and “irresponsible” while denying earlier statements he had made about Sterling’s race) was just a middleman between Merlin and Bob S who helped prepare the letters Merlin would send out to Iran.

MacMahon: You, you told the FBI that Sterling merely acted as a middleman — and this is in 2006 — as a middleman between you and Bob to prepare letters to be included in the package of technical documents, right?

Merlin: Some kind of, yes?

MacMahon: So the person that was making the final say as to what went in any letter you sent as far as you knew was Bob, right?

Merlin: I, I don’t know what is hierarchical.

I raise these comments — both apparently made only after the publication of Risen’s book — because of some oddities in the CIA cables documenting the operation.

Bob S’ 70%

To some degree, the cables that cover the period when Sterling handled Merlin do make it clear the degree to which Bob S was running this operation, and Sterling was just holding Merlin’s hand as he tried to reach out to Iranians.

Over the period in question (the first meeting when Sterling met alone with Merlin was January 12, 1999; he handed over Merlin to Stephen Y on May 24, 2000 (though it appears Bob S had already excluded Sterling from at least one meeting, as noted below), most of the cables written by Sterling deal with the tedium of Merlin’s pay and include — always verbatim — Merlin’s correspondence with the Iranians. Sterling’s cables often ask for input from Langley on Merlin’s drafts; he expresses some concern about the lag during spring 1999 when CIA was getting export control approval for the program.

Then, in the May 13, 1999 cable (Exhibit 24), as Merlin seems to be getting more interest from Iranian Institution 4 (in spite of his having sent his resume and business proposition letter separately), Sterling notes that Bob S will need to inform Merlin where the program heads from here. “[M] should expect a visit from Mr. S who will provide an update on the definite direction of the project. [M] understands that there are aspects of the project that require certain approvals beyond the purview of C/O.”

The next cable (Exhibit 25) describes the May 25, 1999 meeting at which Bob S, with Sterling in attendance, told Merlin that the target of this operation would be Iranian Subject 1. This plan actually dated back to December 18, 1998 (Exhibit 16). In that cable, Bob S referenced a November 20, 1998 cable (not included as an exhibit nor apparently turned over to FBI as evidence) that apparently described IS1’s “new public position” for which he would be “arriving in Vienna in Mid-December to assume his new duties” (one of Bob S’ later cables would identify IS1 as the Mission Manager in Vienna). But it wasn’t until May of the following year when Bob S (and not Sterling) instructed Merlin that he should start finding ways to reach out to IS1. Note, one paragraph of that cable — following on a discussion of IS1 — is redacted.

At the next meeting — on June 17, 1999 (Exhibit 27) — Merlin told Sterling that he was having problems locating IS1, though some of this discussion is redacted.

Then, in spite of the indication that Sterling had tentatively scheduled a meeting for July 5, 1999, we see no further meeting reports until November 5, 1999. (Though on July 23, 1999, someone applied for reauthorization to use Merlin as an asset; Exhibit 29.) It appears that only one cable from this period, which would have been numbered C2975-2976, was turned over during the investigation but not entered into evidence, if the Bates numbers on the cables are any indication. Given the report in the 11/5/1999 cable that Merlin had gone AWOL, it’s likely things were already going south between him and Sterling. From that period forward, Bob S either soloed or attended most meetings with Sterling and Merlin, with one very notable exception.

The exception was the January 10, 2000 meeting (Exhibit 35) at which Sterling informed Merlin CIA would withhold money Merlin believed — rightly, it appears — he was owed. Given that Sterling had already (on November 18, 1999) unsuccessfully requested a transfer out of NY, where he believed he was being harassed for his race, it’s hard not to wonder whether they deliberately sent Sterling out to deliver the bad news, anticipating they’d soon be giving Merlin a new case officer within short order anyway.

All of that is to say that, in spite of the several ways that Sterling appears to have managed Merlin with more professionalism than his prior case officer and arguably even than Bob S, Bob S was running the show, which includes making key decisions and at key moments, dictating how the reporting on the operation appeared.

Two versions of November 18, 1999

To see how this manifested, it’s worth comparing the two cables recording (in part or in whole) the November 18, 1999 meeting between Bob S, Sterling, and Merlin.

The first version (Exhibit 31), written on November 24 by Bob S from Langley and addressed to NY and Vienna — Office #5 — for information, appears under the heading “Iranian Subject 1 is in Vienna” and references a cable from Vienna (this cable, too, appears not to have been turned over as evidence). As such, the cable describes the results of the meeting with Merlin in context of the arrival of IS1 in Vienna, using the “good news” offered by Merlin as an opportunity to flesh out the plan for the blueprint hand off in Vienna. Presumably, paragraph 2 of the cable (which is redacted) lays out the news on IS1’s presence in Vienna. Bob S then presents all the good news involving Merlin in that context with a flourish.

During an 18 November Meeting with [M] Officer [Jeffrey Sterling] and HQS CPD Officer [Mr. S.], [M] provided two pieces of good news. The first was that he has obtained a new [Country A] passport (which he showed C/O’s) and will soon apply for an Austrian visa. His possession of a Green Card should facilitate the issuance of the latter. The second and more significant development was an e-mail dated 7 November which [M] had received from [Iranian Institution 1] Professor [Iranian Subject 2 IS2). [IS2] said he had been going through old e-mailsl and found a 1998 message from [M]. He asked [M] to respond and provide more information about himself. [M] did so in a generic fashion. This contact from [IS2] provides an excellent opportunity to ease [M]’s (and his disinformation packet’s) way in to [Iranian Subject 1 (IS1)] who until recently was also [at Iranian Institution 1] and is still featured on its website.

He then goes on to lay out what he presents as a plan crafted with the help of folks at HQ and Sterling (remember, this was written from Langley, not NY). That plan includes recognition that Merlin is “no one’s idea of a clandestine operative;” to compensate for that, Bob S envisions (resources willing) a Sterling trip to Vienna so he can help provide clear instructions to Merlin as well as Mrs. Merlin traveling to Vienna with the scientist because she was instrumental in his cooperation with the CIA in the first place and is a calming influence.

4) Shortly before he prepares to launch in Vienna (see below RE timing and mechanics) we will have [M] advise [IS2] via e-mail that he is going on vacation in Vienna with his wife and will stop by the Iranian IAEA Mission there with a packet of interesting information for [IS2], asking IS2 to alert the mission to expect [M]. When he shows up at the mission, [M] will have the packet containing the [CP1] disinformation in an envelope addressed to [IS2] and will ask to see [IS1] to make sure the package gets delivered to the right man. [IS1] is likely to acknowledge that he too is from [Iranian Institution 1] and that he knows [IS2]. This will let [M] plant his story (of repeated efforts to find a receptive audience in Iran) more firmly and give the Iranians a chance to see that [M] is indeed a Russian and a nuclear weapons veteran. Even if [IS1] does not see [M] presenting a package with a known addressee at a prestigious Iranian [redacted] institution can only help advance our plan to have the information taken seriously.

5) Per discussion at HQS and with [Sterling], we believe it best to send [M] to Vienna with his wife in early January (after the Austrian Christmas pause and the Islamic holiday of Ramadan, which begins on 9 December and ends on 8 January) to make the approach to [IS1]. His wife, [Mrs M], was instrumental in getting him to cooperate with [CIA] in the first place and is a definite calming influence on him. [M] is no one’s idea of a clandestine operative and we believe it wiser to refrain from meeting him while he is in Vienna. That said, he needs to be thoroughly prepared. One option – contingent on available resources – would be for [Mr S] and [Sterling to] visit Vienna during the first week of the New Year [redacted] so he can given the rather differently-oriented [M] as much concrete detail about where he has to go and what he has to do as possible. [1 line redacted]

Spoiler alert: while Mrs. Merlin did travel to Vienna with her husband (and probably had a big role in even getting him to go and — my suspicion is — had a role in the operational security measures Merlin took which helped doom the operation, though neither she nor the CIA would ever admit that), Sterling never did make the trip, and Bob S’ instructions — which Bob S’ habit of flourish aside were probably also deficient because he was too familiar with the city — ended up being one of the problems with the trip. It’s worth mentioning, too, that according to Bob S’ testimony, he made several trips to case out Iran’s IAEA mission in the months leading up to the operation and one of his cables describes having done so too.

Now compare Bob S’ cable with Sterling’s (Exhibit 31), written on December 1, 1999, a week after Bob S’ cable and 12 days after the actual meeting (it’s probably worth noting that on the very same day this meeting took place, Sterling asked for a transfer out of CIA’s New York office, and within 5 days his boss was scolding him for having done so), and addressed to Langley and — like Bob S’ cable — Vienna, for information.

Sterling saves his enthusiasm over the outreach to Merlin from IS2 for his last paragraph.

Feel this is a fortuitous turn of events for the operation, as a preliminary thought, the contact from [IS2] can be exploited to either provide another person to present the material to, or somehow utilize this contact to provide a more definite entree to [IS1] for [M].

Curiously, that paragraph seemed to show little awareness of Bob S’ extensive plans for how to exploit the IS2 contact to provide “a more definite entree to IS1,” even though Sterling references the cable Bob S wrote.

Aside from the first, action, paragraph in Sterling’s cable (which is redacted), the sole apparent explanation for why he wrote a cable after Bob S had already written one reporting all the same news from the meeting as Sterling would seems to be the inclusion of the verbatim content of the outreach from IS2.

During the meeting, [M] mentioned that he had received the following email from [Iranian Subject 2 (IS2)] from [Iranian Institution 1] dated 7 Nov:

Dear [M]

I was reviewing my old mails. I found you last year email. I want to know more about you. Could you let me have more information regarding your work, your hobby, your interest, etc?

Regards,

[Iranian Subject 2]

[IS2]’s email address is [redacted]

It’s not surprising Sterling included the verbatim email — he always did that in cables he wrote solo. It’s just rather curious that Sterling submitted his “preliminary thoughts” — along with the verbatim language — so long after Bob S had rolled out his plan.

Prelude to a clusterfuck

The next cable (Exhibit 33), dated December 16, 1999 and describing the December 14 meeting between Sterling, Merlin, and Bob S, reflects continued uncertainty about how to get Merlin to Vienna in such a way that he didn’t screw up the operation. “[M] has and will be provided with enough information so that any concerns he will have about finding the building should be alleviated,” the cable optimistically predicted. At that point, however, it wasn’t getting lost that had Merlin worried. It was that his wife would find out what he had been up to (though she almost certainly already knew).

When asked, [M] expressed as his main concern actually carrying the documents on his person when he travels to Vienna. [M]’s preference is that his wife ([Mrs. M]) not know any specifics about his work for the CIA. He feels certain that she will discover the package and have many questions that he would prefer not to have to answer.

Note that the action paragraph of this cable is redacted.

By the following meeting, the ill-fated January 10, 2000 meeting documented in a January 12 cable (Exhibit 35), however, Merlin had resolved these concerns. When Sterling said that CIA would arrange to have someone meet him with the blueprints, Merlin explained that was no longer necessary.

[M] said that the situation has changed and that he can now take the package. [M] explained to his wife that he has to deliver some materials while they are in Vienna. He did not give her any further explanation. [M] said the reason he decided to tell his wife was that he thought it might be too risky to have someone meet him in Vienna, so he felt it more secure to handle the package himself. [M] said the he has not apprehensions about being in Vienna alone, but that he would like to have an emergency contact number just in case.

At the same meeting, Merlin and Sterling discussed two other aspects of the drop-off that would be significant. First, that “the best way will be for [M] to simply drop the package off and then depart the mission without any lengthy discussion with Iranian officials.” And also, “the possibility of two letters being included with the package,” the second of which would be a hand-written note to IS1 telling him how important the materials in the main package, addressed to IS2, were.

So after having talked to his wife about delivering sensitive materials in Vienna — but he didn’t provide any more details, promise! — the following discussions, which would each contribute to the Vienna clusterfuck, began:

• Merlin would carry the packet; in the process, he would take out (and, according to one of his stories, never put back in) the names of certain nuclear devices
• Because he carried the packet, he was able to bring the letter on disk (hidden among 19 other disks), meaning he could destroy the disk without ever giving CIA a final version of what he included
• Merlin would bring Bob S’ cell phone number as an emergency backup, which he would use to place a call from his hotel phone, only to be instructed to follow the directions he already had
• Merlin may have taken the permission to simply drop off the package without lengthy discussion and turned it into dropping off the package with no discussion
• Merlin combined that permission to just drop the document with the discussion of a second, hand-written note to justify leaving a significantly different hand-written note, the content of which the CIA also cannot be sure of

The only other step Merlin is known to have taken that screwed up this operation — refusing to leave his PO Box for follow-up content — had been in the works since the previous summer, when he had started to do the same with those letters.

In other words, Merlin talked to his wife — but he didn’t provide any details! — and then proceeded to implement the steps that in his mind he needed to do to protect himself and his family even while potentially implicating the CIA directly in the drop-off, while ruining several of the operational goals for this operation.

And he proceeded from adopting those steps to launching the first of two refusals to do the operation without getting paid more (in each instance, Merlin would apologize via phone the following day and say he would do the operation, a capitulation Bob S attributed to Merlin’s wife in his testimony).

Also in this meeting — which took place 7 weeks before Merlin left for Vienna — Sterling and Merlin worked on the fifth iteration of the letter, which Sterling included verbatim in this cable. I’ll return to that in a follow-up post.

Because Merlin launched his payment strike (for which Bob S apologized to Sterling in a January 14, 2000 cable), Bob S was forced to come to NY to try to appease Merlin. He did so for a February 17, 2000 meeting, detailed in a February 17, 2000 cable (Exhibit 37) written from Langley and addressed to NY, Vienna for Info, and CIA offices 7 and 8 (the liaison services of which CIA would later ask to track any signs of response from Iran; because of that one is likely to be Tel Aviv). In that meeting, too, Merlin walked out because of his payment dispute. In that cable, Bob S described what would happen if, on a follow-up visit, he deemed Merlin prepared for the operation. It includes the instruction that “we will need a full and detailed report of his visit and reception.” Bob S’ cable documenting that February 21 follow-up meeting — a February 22, 2000 cable (Exhibit 38) — described his judgment (Sterling did not attend this last meeting) that Merlin was prepared. “[Merlin] and the information he is carrying have been exhaustively prepared,” Bob S alone judged, “and now it is up to luck and the Iranian reaction.”

Bob S’ two version of Mission Accomplished

I’ve already written about the two different versions of Mission Accomplished cables that Bob sent and will write at more length in a follow-up post. For the purposes of this post, however, it’s important that Sterling wrote neither of them even though he attended the March 9, 2000 meeting at which Merlin described his trip, though Bob S claimed in his testimony that Sterling “may have been sitting at the next terminal” when he wrote the first of them. For the purposes of this post, however, it’s worth noting what Bob S did and did not include.

Bob S’ March 10, 2000 cable addressed internally (Exhibit 44) did admit:

• Merlin called Bob S’ cell phone from his hotel room phone
• Bob S’ details included some errors (here described as mis-counting the number of steps leading into the Iranian mission building)
• Merlin showed up one day while people were in the mission without going inside (purportedly because he didn’t have the packet)
• Merlin left the packet without speaking to anyone
• Merlin took photographs of the mission

But the unredacted parts of that cable (the action paragraph and one more are redacted) did not admit details that are now part of CIA’s operative story (though may not all be true):

• CIA reportedly has no record of what Merlin left, neither in the computer printed nor the hand-written note, in part because he destroyed the disk on which he had written the former
• Merlin did not include his PO Box, as instructed, for further contact
• Merlin may have substituted “Device 1” for the actual names for key devices in the schemes
• Merlin did not write a report, as instructed

Bob S’ March 13, 200 cable (Exhibit 3) included a tearline intended for “local intelligence services” in 3 overseas locations (which I take to mean Israel’s and two other countries’ intelligence services were the target audience). In addition to the other things Bob S suppressed in his March 10 cable, he:

• Falsely implied Merlin had included his PO Box for further contact
• Hid that Merlin had been at the Iranian mission on one day when people were present
• Made no mention Merlin had called Bob S from his hotel room
• Made no mention of Merlin’s claimed difficulties finding the mission
• Hid that there were two separate letters — the  handwritten one and the computer print out one

The point, of course, is that in cables Bob S wrote immediately after debriefing Merlin after the operation, he was being less than fully truthful, both to liaison partners, but even for internal reporting, about a number of the ways that Merlin had blown off his instructions.

Bypassing Sterling

Then there’s the most curious cable from the consideration of Bob S running an operation on which Sterling was just a (per Merlin) “middleman.” On May 24, 2000, Sterling handed over managing Merlin to Stephen Y (Exhibit 47). Before then, on April 5, 2000 (Exhibit 45), Merlin “was met” (note the passive voice, which seems to violate CIA’s protocol for cable writing, which puts the details about meeting attendees in the second paragraph) to see if he was interested in participating in a similar operation, only targeting a different country which was almost certainly Iraq. The cable — written in NY, addressed to Langley (for information) as well as Vienna and the same three CIA offices where the CIA was seeking liaison help tracking the Iranian op, and apparently written in Bob S’ fluffy style — describes Bob S making the ask. But then it describes case officers, plural, being “glad that [Merlin] posed no objection to a rather more adventurous extension of the current operation.”

By all appearances, even before Sterling handed over Merlin to his successor, Bob S was holding meetings with Merlin without Sterling’s involvement (and this is consistent with trial testimony that seemed to suggest that Sterling would suspect but not know of the other countries involved, as indicated by Risen’s book).

Bob S got Sterling to hold Merlin’s hand through a disastrous delivery of one set of nuclear blueprints, and even though Bob S admitted — in a highly self-serving cable — Merlin’s “inability to follow even the simplest and most explicit direction,” Bob S was asking Merlin, outside Sterling’s presence, to approach (probably) Iraq a month later.

The late admission of Bob S’ 70% thinking

After Merlin’s 2003 interview with the FBI, he told them he would tell Bob S if he remembered any other details about Sterling. Bob S was still managing the Merlins in 2006 when he met with them twice about Merlin’s book. And in 2006 — but not, apparently, in 2003 or 2010, when both had at least one other interview with the FBI — Bob S and Merlin were both telling a story about how minor Sterling’s role in the operation was.

Both denied having done so in their sworn trial testimony.

Perhaps they did so — Bob S did so — because of his fairly transparent efforts to include others in any blame for this clusterfuck, implicating both Sterling and the “Generals” he said who had approved every step of the operation, in his extended effort to use the trial to prove this wasn’t a clusterfuck.

But the claims, in 2006, that Sterling wasn’t all that involved make me wonder whether Bob S was prepping a claim that Sterling wouldn’t know precisely what the operation was about given that he was doing just 30% if the thinking on the operation.

Update

First, here’s the working document I used for this post. In includes three things: A side-by-side comparison of the two cables describing the November 18, 1999 meeting, a side-by-side comparison of Bob S’ two Mission Accomplished cables, and a list of all the cables from when Sterling managed Merlin. As part of the latter, I tracked the Bates numbering of cables. Each cable should have a Secret cover-sheet not included. Thus, I surmise that any 4-Bates number gap includes a 3-page cable (cover sheet plus two pages of content) plus the cover sheet for the next cable. The most significant detail from the Bates numbering is that the second Mission Accomplished cable comes from a different part of what appears to be CIA production (C115-116 as opposed to C2991-2992). That may mean it was found in someone else’s hard copy collection; the rest likely come from Bob S or CPD, though the cable gap in the series may reflect that same cable. There’s likely nothing interesting in the missing cables; after all, if there were something interesting, the defense could have submitted it, as they did the second Mission Accomplished cable.

Second, there’s a line that has stuck in my mind since writing this post. In Bob S’ cable describing the last meeting with Merlin before the Vienna trip (Exhibit 38), he writes, “Perhaps characteristically, [M] had misplaced the e-mail address of [Iranian Subject 2] and [Bob S] provided it again along with instructions to send off a brief notice telling [IS2] of his plans to deliver an important packet to the mission in Vienna.” First, while Merlin was flaky about a lot of things, there’s no evidence he was flaky about losing emails (though this may have reflected Merlin’s efforts to avoid more personalized contact). Also note Bob S says he “provided” the email “again.” I can’t think of when he would have provided it before (unless it was back in 1998), at least per the operative story.

Then, in his first Mission Accomplished email (Exhibit 44), Bob S says he and Sterling “directed him to send a follow-up e-mail to [IS2 at Iranian Institution 1] informing him that he had dropped off an important packet of information in Vienna and asking [IS2] to confirm its receipt.” In his second Mission Accomplished email (Exhibit 3), Bob S claims “the asset e-mailed the professor before and after the Vienna trip to alert him to expect a packet of valuable information.” The thing is, because Sterling (who was very good about recording such things) stopped writing the cables, we have no way of knowing whether Merlin ever got a response from IS2 after his “generic” response to IS2’s initial November 7 email before the November 18 meeting. A January 14 Bob S cable (Exhibit 36) reflects the instructions that Merlin should send both an email and a letter, but there’s no record they reminded Merlin of that at the February 14 meeting and Bob S had to reiterate the instruction to send an email at his February 21 meeting. And there’s no indication that Merlin had sent one between the March 10 and March 13 cables. In other words, we have no cable record of Merlin having emailed before and after, as Bob S claims in his cable. Thus, it’s possible the tie with IS2 was even sketchier than it seems (and certainly, Merlin never got any confirmation from IS2, which suggests he never heard back from him).

Finally, particularly given his varying claims about Sterling’s actions, it’s worth noting two aspects of Bob S’ relationship with Sterling. First, there’s a dispute about what Bob S said when he took Sterling aside to deal with the concerns Sterling raised about Merlin’s initial reaction to the nuclear blueprint. In his testimony, Bob S said Sterling was “taken aback” by Merlin’s response. But after much effort to deny it, Merlin testified that Bob S “did tell Jeff to shut up in this discussion.” Nevertheless, when asked this on cross, Bob S specifically denied “telling him to shut up.”

Then, during cross-examination but in response to questions from Judge Brinkema, Bob S admitted that Sterling had told him “a handful” of times in 1999 that he had been treated unfairly because of his race. In response to Sterling telling him of this, Bob S told Sterling “he needed to do his job and not worry about it.” This almost certainly would have been around the time of the November 18, 1999 meeting. None of that means Bob S had it in for Sterling. But it does suggest he was entirely unsympathetic to both his operational and professional concerns.

 Update

On review I realize Merlin told the FBI in 2006 — the same year both he and Bob S said Sterling was more tangential to this operation — that “the details of this operation were a wild forest to Sterling.”

I’ve finally gotten a hold of the transcript for Merlin’s testimony in the Jeffrey Sterling trial (working on getting something I can post; he was apparently difficult to understand, in any case, so not even people present understood all this).

Reading it, it’s clear why the government has claimed, going back to 2011, that Merlin’s imminent death from cancer meant he should not testify. I don’t dismiss the gravity of his health problems (and also note that he is apparently on pain killers, including Oxycontin, which may have affected his testimony here). But he was a terrible witness, and pretty clearly lying on a great number of accounts.

But I’m interested in specifically how he denied things that appear either in James Risen’s book or in CIA cables.

It’s lie

About two things, Merlin was adamant. The first is the same thing that really elicited the Merlins’ ire when they read Risen’s book: the report that they were defectors.

Trump: It says you defected to the United States. Is that accurate?

Merlin: It’s lie.

Note, given the timing and the claim that Merlin might have been involved with the Soviet Union’s 1980s-era nukes, I entertain the possibility that they defected to some other country before moving to the US in the early 1990s. That’s true, especially, because when Merlin got his passport renewed in 1999, he did so from a country the name of which got substituted (meaning it probably wasn’t Russia; the original appears to be 9 characters long, so Ukrainian is a possibility), though it could just be a successor state. Whatever the case, the timing of the Merlins’ arrival in the US and their certainty with which the government repeatedly said they did not defect convince me that Merlin is correct here: they were not defectors.

Similarly, Merlin is equally adamant that the description in Risen’s book that Merlin tried to warn the Iranians of “flaws” in the blueprints he handed them was not true.

Trump: In paragraph 64, the book represents on page 205 that the letter was warning the Iranians as carefully as you could that there was a flaw somewhere in the blueprints. … Was that the purpose of the letter?

Merlin: It’s, it’s lie. [Later] I don’t see flaws here. It was just incomplete information.

While it’s certainly true that Merlin’s and the government’s understanding of the significance of the incomplete information in the blueprints was very different — elsewhere Merlin claimed that a real fireset schematic was “100 times more complicated than it was shown in drawing and the schematics” — it is also true that Merlin appears not to have known about the deliberate flaws US scientists put in the blueprints. So he is correct that the representation in Risen’s book is incorrect on that point.

I don’t remember

Then there are a series of questions about which Merlin likely feels some shame, about which he professed not to remember the correct answer. One of those topics pertained to whether his wife also spied (note, Merlin and the CIA both are almost certainly lying about how much Mrs. Merlin knew about this operation).

Trump: Did your wife at the time also agree to cooperate with the CIA?

Merlin: No.

Trump: Did she eventually?

Merlin: She didn’t know anything about it.

Trump: She didn’t know anything about what you did, is that correct?

Merlin: Yes.

Trump: But she was interviewed from time to time by the CIA as well?

Merlin: I don’t remember. Probably.

Merlin’s wife remained on the CIA payroll after he claims he stopped getting paid. Surely he knows that. But he’d prefer not to admit it.

Another of the topics about which Merlin forgot the correct answer came in response to a defense question about whether he ever used his American PO Box in communications with Iranians.

MacMahon: Did you testify earlier today that in all of your communications with the people, the Iranian institutions or otherwise, that you, you didn’t use any kind of an American address in any of those documents?

Merlin: I don’t remember.

Now, it’s possible Merlin’s earlier answer on whether he had used his PO Box on correspondence with Iran is correct: that is, it may be that he always ignored CIA’s orders to do so, and CIA simply never found out about it (perhaps in part because the case officer before Sterling did not track that correspondence as closely as Sterling did). But the CIA record shows that he first started balking about using his actual geographic location about a year before going to Vienna, but before that had publicly used his PO Box.

I don’t know

Then there are a series of questions where Merlin clearly either had forgotten key details, or wanted to avoid admitting the truth.  For example, when asked by prosecutor Jim Trump (who had met with Merlin before this deposition to go over it) whether this was a rogue operation, Merlin first offered up that it was a “brilliant” operation (elsewhere he took credit for Iran not have gotten nukes since 2000).  But when asked a question to which the answer is clearly yes — whether it took significant persuading for Merlin to complete this operation — he claimed he didn’t know.

Trump: It states that prior — prior to your trip to Vienna now is what is being discussed here. “It had taken a lot of persuading by his CIA case officer to convince him to go through what appeared to be a rogue operation.” Is that accurate?

Merlin: It was not rogue operation at all. It was brilliant, brilliant operation.

Trump: Did it take a lot of persuading by you — excuse me, by your case officer to go through with the operation?

Merlin: I don’t know.

Merlin walked out of the meeting on final preparations, after having walked out of the meeting prior. That wasn’t, apparently, because Merlin cared whether this was rogue or not, but because he thought the risk to him was too great for the money he was being paid. But the answer to whether it did take persuading should have been yes.

Just as interesting, when Merlin was asked by defense attorney Edward MacMahon whether he had ever before this deposition told the FBI or CIA he had destroyed the disk on which the final version of the letter to the Iranians, he said he didn’t know.

MacMahon: The first time you–you were, you were asked questions over, over a space of many years, and you never told the FBI at all that you had destroyed the disk that you took to Vienna, did you?

Merlin: I don’t know, but there was, was no reason to bring it back. It just put myself in additional danger to have such disk in possession. If somebody stop me and read this disk, I’m in trouble.

MacMahon: Okay. But you didn’t tell the FBI, you didn’t tell anybody until today as a matter of fact that that’s what your story was as to what you did with the disk in Vienna, correct?

Merlin: I don’t know, but again, it was no reason to keep this disk when action was, operation was accomplished, and no reason to keep it as a drawing, as letter, as whatever.

The answer is clearly no, but Merlin doesn’t want to admit that for some reason (I’ll return to the significance of this question in a future post).

In a related vein, Merlin went to some lengths to avoid confirming some things he had told Agent Hunt in 2003 (importantly, before anyone knew what Risen would eventually write about the Vienna trip, including that Merlin wrote his own letter at the end). He professed not to know that he wrote letter in defiance of his orders from Bob S (he had, in fact, discussed doing such a thing with Sterling, but as Merlin confirmed elsewhere, Bob was in charge of the operation), and in doing so, provided the Iranians specific information about what they were getting in the blueprints.

MacMahon: Do you remember in 2003 telling the FBI, Agent Hunt specifically, that the recipient, that the — and I’ll read it to you: “When asked if he could recall the content of the note, he advised that he informed the recipient that the enclosed material was 90 percent complete.”

Merlin: I don’t know. It’s strange.

MacMahon: You don’t remember that?

Merlin: I don’t know.

[snip]

MacMahon: Did Bob tell you to deliver a handwritten note with the package?

Merlin: No.

MacMahon: And Jeff Sterling didn’t tell you to do that, either, right?

Merlin: No.

MacMahon: And did that note say — was it the handwritten note — do you remember the content of the handwritten note?

Merlin: I told you already it was just statement: “I couldn’t reach you. Nobody was there. I’m leaving the package. There’s valuable information in it.”

MacMahon: Okay. And so when it said — when you told the FBI in July of 2003 that part of the content of the note was that the enclosed material was 90 percent complete, that wasn’t in the handwritten note, either, right?

Merlin: I don’t know what you’re talking about.

I had to protect my family

That Merlin is not remembering inconsistencies with his past admissions that go to the core of whether Risen’s book is largely on point about the clusterfuck of the operation is particularly interesting when, presented with the way in which some of these very same actions diverged from instructions in other questions, Merlin aggressively defended them as necessary to protect his family.

MacMahon: What you wanted to do was to leave an e-mail address, correct?

Merlin: Yeah.

MacMahon: And that’s what you did when you were in Vienna. You left a note that contained an e-mail address but didn’t attach any kind of contact information for you personally in the United States, correct?

Merlin: Correct.

MacMahon: All right. And that was not what you were told to do, was it?

Merlin: But it was my protection.

MacMahon: But it wasn’t what you were told to do, correct?

Merlin: I would say yes.

[snip]

MacMahon: Do you remember being told to put an address for your — in the United States for yourself in the package that you delivered in Vienna?

Merlin: I believe it could be more trustful if I represented like Russian scientist than scientist living, living in the United States. Nobody likes United States in the world.

[snip]

Mac: Did–before you left for Vienna, your last meeting was with Bob, wasn’t it?

Merlin: I don’t remember. I believe it was Jeff.

[snip]

Mac: Is it your recollection that any of those meetings, that you were told by either Bob or Jeff, Jeffrey Sterling, not to include an American address in your letter that you were going to give to the Iranians with the pans for a nuclear weapon?

Merlin: I offer it, but I’m not stupid. I can put in danger my family.

It’s the same answer that Merlin dodged elsewhere — that he deliberately ignored the clear instructions Bob S had given (here, as elsewhere, he falsely blames Sterling for stuff the cable record clearly shows Bob S did, but I suspect that is the way he remembers instructions he hated). These are precisely the actions Merlin took that made this operation a clusterfuck. But whereas when asked in the context of whether that made the operation a failure (and the book an accurate portrayal of why it was a failure), Merlin claimed not to remember, his memory instantly and aggressively returned when presented an opportunity to defend his own actions from an operational security standpoint.

As I noted here, the cable record suggests that when Merlin recognized how the CIA’s efforts to dangle him left his identity and location exposed, he started taking his own measures to ensure his own safety (I wouldn’t be surprised if that change in behavior came with the disclosure to his wife, or her discovery, of what he was up to). He’s not ashamed or forgetful about those measures in the least.

He just wants to hide how directly those actions led to this operation being a laughable failure.

Bob S — the guy in charge of the Merlin operation — appears to have made a false claim under oath at the Jeffrey Sterling trial.

To be fair, I think he told a lot of fibs — shading his description of a program that was an operational disaster as something less laughable, even as walking the court through documents that (upon a close read) make it clear how laughable the program was. And I suspect — though cannot yet prove — that he engaged in far more serious deceit while testifying against Jeffrey Sterling under oath.

But according to CIA’s own records, Bob S was not telling the truth when he claimed, repeatedly, “our record keeping was better than” Merlin’s about his complaint, in early 1999, that he hadn’t been paid his full salary the previous year was incorrect. He was not being fair to Merlin when he claimed, when Merlin complained again in early 2000 — just weeks before Merlin would travel to Vienna to hand the Iranians a newspaper-wrapped nuclear blueprint — that was “getting cocky” when he demanded he get paid his salary.

To be sure, Merlin also appears to have bilked the CIA over the years, charging them for things like his long distance phone line and a $200 modem, presumably as a way to increase his take from the Agency.

But the CIA knew Merlin was working with them for the money. “[M]’s operational motivation for this activity is almost purely financial., and his desire to continue earning income from [CIA] something that [M] was very frank and honest about during his meeting with C/O and CPD officers,” his then-handler Laurie D wrote in a cable (Exhibit 5) pitching him for the operation in January 1997.

In spite of the importance of money to Merlin’s motivation, they do appear to have botched their record keeping, every bit as much as Merlin. And it was an issue that — even according to Bob S — poisoned the working relationship between Sterling and Merlin. But it also appears to have made Merlin much less willing to do what CIA wanted him to do.

On February 24, 1999 (Exhibit 21), Sterling and Merlin met for 3 hours. During the meeting, Sterling paid Merlin “one-third of his ’98 salary (USD 20,000.00)” but then reminded Merlin that he could only be reimbursed for things tied to his work with the CIA and expressed confusion that Merlin was billing the CIA for two phone lines when he should only be billing for the one tied to Internet access (by this point, Hotmail). “From that moment on,” Sterling told Merlin, “[M] would only be reimbursed for the phone line dedicated for use by [M] in furtherance of the project.”

Merlin responded by noting he had been making the same amount (so, presumably $60,000, though Langley disputed that in his testimony) for two years, and “some adjustment may be in order.” Merlin suggested that a raise would do away with his need to charge the CIA for petty expenses like the phone line. While an entire paragraph of this discussion is redacted, in response, Sterling and the New York office recommended he get at least a $250/month raise, and asked Langley to consider Merlin’s request for life insurance, given that he was contemplating meeting Iran, which raised some risk.

On March 16, 1999, Sterling explained (Exhibit 22) that Merlin’s new agreement “will include an increase of USD 1000.00  in his monthly salary. [Sterling] said that with such a sizable raise, there would be some changes with regard to the reimbursements that [M] has been receiving. [Sterling] explained that [M] would not longer be reimbursed for phone expenses related to his use of the Internet for the project.” And Merlin was on his own for life insurance. Again, almost a paragraph of this financial discussion is redacted, but that should have raised Merlin’s salary to either $72,000 (if the $60,000 Sterling used) or $67,000 (using the disputed amount) once he signed his new agreement.

On April 12, 1999 (Exhibit 23) — less than two months after handing over “one-third of his ’98 salary (USD 20,000.00),” Sterling “had to inform [M] that the balance of his 1998 salary is USD 55,000.00 as opposed to 60,000.00.” So Merlin did what any underappreciated worker might do. “[M] threatened to quit since the money discrepancies seem to crop up every year. [Sterling] immediately challenged [M] on his statement asking [M] if he was ready to quit based on a mere USD 5,000.00 (especially in light of the total amount of his salary). [Sterling] requested a point blank answer from [M] on what his actions will be. [M] calmed down and said that he is weary of the same pay discrepancies occurring year after year, but said he will not quit the project.”

Sterling went on to suggest he, as the case officer, might not have been entirely certain what was going on. “[Sterling] told [M] that the reason for the discrepancy will be found, and that it is not beyond the realm of possibility that [M] has already been paid for one month out of ’98 that would make his ’98 total USD 55,000.00. [M] then signed a receipt for USD 35,000.00. To date, [Sterling] has paid [M] USD 55,000.00 representing his 1998 salary.”

Most of the following paragraph is redacted, though it speaks “also” of a “M/2’s” status update — perhaps a reference to Merlin’s wife, who was also a CIA asset.

On May 5, 1999, Sterling seemed to confirm that Merlin had been correct (Sterling appears to have used the wrong date in paragraph 2). (Exhibit 24) “[Sterling] paid [M] USD 5,000.00 as the balance of his 1998 salary (USD 60,000.00). [Sterling] believes that the previous confusion with regard to the amount of [M]’s 1998 salary was based on the fact that his [agreement] spans a Feb – Jan timeframe which is somewhat different from the normal fiscal year timeframe of Jan – Dec.”

There’s a big jump in cables before the next meetings described, in part because the government didn’t introduce the July to October ones into evidence, in part because Merlin went AWOL for the month leading up to a November 4, 1999 meeting. From what we see of cables between November and January, though, money issues don’t arise again until January 10, 2000, at one of what appears to be the first meetings in a while where Sterling meets with Merlin without Bob S as well. As Sterling notes in the summary of the meeting (Exhibit 35), “Despite the progress made and [M]’s apparent readiness, issues related to [M]’s salary [redacted] have placed doubt as to whether [M] is willing to continue with the project.” Here’s what Sterling had to tell Merlin just weeks before he was supposed to deliver a nuclear blueprint to Iran in Vienna:

[Sterling] took pains to explain to [M] in a reasonable fashion that the current payment scheme was causing problems and that a new structure had to be introduced. [Sterling] also explained that [M] would be receiving a [“additional information regarding his 1999 salary” replaces 3-4 redacted lines]. [M] had no difficulty that his future salary would be paid to him as earned [one line redacted] [Sterling then told [M] that as a result of the measures being taken to correct his salary situation, review of his salary history indicated that he had been overpaid by USD 5,000.oo in Feb ’98. And, as a result, his ’99 salary would have to be reduced by 5,000.00. [Sterling] had 60,000.00 for [M] representing Feb -Nov ’99 (6,000 per month). Though [M] earned 66,000 for Dec ’99, this amount was reduced by 5,000.00 per HQS information that [M] was overpaid by 5,000.00 in Feb ’98. [Sterling] chose not to bring the remaining 1,000.00 in anticipation that [M] would not understand the reduction in his salary amount. During the conversation, [Sterling] tried to explain that the remaining amount (either USD 1,000.00 or 6,000.00) would be paid at the next meeting once it is clearly determined that [M] had in fact been overpaid by 5,000.00.

[M became incensed and said that [Sterling’s] infoformation was not correct. [M] said that the money he received in Feb ’98 was for a Dec ’97 payment that he had not received and therefore had not been overpaid in 1998 as [CIA] contends. [M said that he has waited too long for his finances to be corected and that he did not wish to  proceed with project any longer. [M] then proceeded to blame [Sterling] for the salary problems. However, [Sterling] quickly reminded [M] that the difficulties experienced with his finances were a result of activities prior to ’99, i.e. before [Sterling] was involved.

After some more back-and-forth, Merlin left the meeting. Sterling called him the following day. And while Merlin was calmer, he still said that “he will not proceed with the project unless and until he receives [additional information about his salary] and USD 66,000.00 that he believes he is due, or a promise from us that these items are coming to him.” In the cable, Sterling and his manager suggested that Bob S travel to NY to explain and resolve it.

Two days later, Bob S wrote back to Sterling (Exhibit 36), apologizing that Sterling had to do the dirty work, but showing little sympathy for Merlin.

HQS regrets that [Sterling] was a victim of the murdered messenger syndrome after bringing (not very) bad news to [M]. Any confusion about [M]’s salary is largely his own fault because he wanted to be paid different parts of his salary in different years. That said, he may be right about the early 1998 payments, and he is evidently quite emotional beneath the stolid surface and not capable of sorting it all out rationally. He has had a lucrative relationship with us since 1994 and is acting in an immature fashion. Nevertheless, we need his services now and [Mr. S] will seek to placate him. We propose paying him the disputed salary. We will carefully consider an appropriate operational bonus upon the successful completion of his Vienna mission.

The next cable, describing a February 14 , 2000 meeting with Merlin (but written by Bob S back to NY), revealed that Merlin remained pissy about the salary issues, actually walking out of the meeting at which Bob S was supposed to placate him. Bob S doesn’t provide much detail on what happened, describing Merlin’s “histrionics” as having to do with “minor proposed changes in his [agmt],” judging that “none of [M]’s desires concerning his [agmt] are show stoppers (cash payment, [redacted] a small disputed sum),” and stating he would bring the previous year’s agreement and work off that to a follow-up.

Bob S’ account of that follow-up visit (which Sterling did not attend; Exhibit 38; this time the cable Bob S writes the cable from NY to Langley) indicates “we could meet him halfway on when and how he is paid,” though the roughly 4-5 line description of what that means is redacted. Bob S “paid him $66,000 for his 1999 earnings and provided $5000 as a travel advance.”

Which is what happened immediately before Merlin got on a plane to Vienna and failed to follow most instructions about how he was supposed to hand over a nuclear blueprint to Vienna.

Now, I lay this out not just because it shows CIA’s dysfunction again. All the more when you compare the numbers submitted as a stipulation at trial (see above), which make it hard to understand how Merlin is not absolutely correct that the CIA stiffed him in 1998. Even if his salary was supposed to be just $55,000 and even accounting for the weird accounting he apparently requested, he received less than $49,000 in a year when 11 months at a $55,000 rate would have been $50,416. And at least from the narrative we have, Sterling made all the payments to Merlin that transpired in 1999, and the money that Bob S ultimately paid Merlin would might work out to be $71,000, but that would not seem to account for his $1,000 raise. It’s possible they changed all that in the redacted bits (or, as CIA seems to like to do, retroactively). But both the confused actions of Sterling and Bob S and the actual numbers compared to the stated numbers in the cables suggests, at best, that CIA’s accounting system is just as screwy as Merlin’s could have been, if not worse. (Note, the first several years of Merlin’s finances with the CIA also don’t appear to match the testimony of his first case officer, Stephen B, who said they had a dispute over whether Merlin would work one year for $150kK or two for $300,000, the latter of which is what the CIA wanted.)

So there was no reason for Bob S to claim that Merlin was in the wrong. At the very least, CIA’s records were so fucked up, neither a case officer or a program manager could figure this out over months and years. And it at least appears Merlin is in the right (but then, what court can he appeal to?).

At one level, I attribute this to just more Bob S spin — along with his inability to hide his disdain for others and his real need to blame others for the clusterfuck that became the Merlin program.

But it’s important to identify because it raises one possible motive for the Merlins to want this story to come out (remember, at first they weren’t all that bugged by the book, until they realized Merlin looked bad in it). But it also puts another perspective year-and-a-half leading up to the beginning of Sterling’s disastrous end with the CIA. They were treating his job as case officer to fix the financial screw-ups made years earlier. Bob S sent him out to do that alone, and only after came in to rain down cash on Merlin.

It was probably just garden variety CIA (even, generic bureaucracy) screw-ups behind the scenario. But nevertheless, it likely had real consequences both for Merlin’s willingness to do his job as ordered and Sterling’s feelings about the trustworthiness of the Agency.

On March 16, 1999, Jeffrey Sterling met with Merlin, the Russian scientist Sterling was trying to get to establish ties with Iran so he could hand off a nuclear blueprint. (Exhibit 22) Merlin seemed to be getting impatient — and perhaps a little insulted — that the Iranians he was approaching weren’t showing more interest in his 20 year experience as a Russian nuclear engineer. So he made an utterly bizarre suggestion.

[M] then suggested that in some of his future messages, he may make mention of the recent revelation that another country had secured nuclear secrets from the U.S. [M]’s reasoning was that others now see that it is possible to obtain nuclear secrets which can advance their programs, and that the project can build upon that supposition to entice the Iranians. [Sterling] lauded [M] for his thinking but said some thought would need to be given to such a proposition prior to [M] implementing it.

Merlin has to be referring to the stories about Wen Ho Lee which started appearing on page one of the NYT starting on March 6, 1999. (Remember, too, that Merlin lived in the NY area, so if he read this in the dead tree version — as most people still read newspapers in 1999 — he most likely read it in the NYT.)

Those stories were written by James Risen.

Which is strong evidence that Merlin was reading Risen as far back as 1999.

Merlin’s suggestion — that he, a CIA asset, entice Iran to accept his Russian blueprint by pointing out that China had allegedly jump-started its own nuclear weapon program by stealing blueprints from the US — reveals just how unclear on the concept of the operation Merlin was. After all, it had to have been suspicious enough to Iran that a Russian who had moved to the US was seeking to deal blueprints (it’s unclear whether the blueprints were ultimately in English or Russian). Any suggestion that the Iranians would thereby be getting US, as opposed to Russian, technology should have alarmed them greatly.

It’s also, of course, a bizarre commentary on the arc of Risen’s career, that the main character in a future book of his would be monitoring nuclear developments by reading Risen. Risen, of course, managed to protect his sources in both cases, in a series that unfairly identified Wen Ho Lee as a Chinese spy and in a book that raised real questions about what our nuclear establishment was doing.

I’m still waiting for Merlin’s transcript on this point, but his wife was asked whether she and her husband knew of or knew Risen. “I start to know about Jim Risen after he wrote the book,” Mrs. Merlin testified on the stand in her imperfect English. She went further, asserting that her husband did not know (it’s unclear whether she meant “of,” or “personally”) before, either. Given how much of the Wen Ho Lee story was driven by Risen between March 6 and March 16, 1999, Merlin probably had known “of” Risen for years before Risen started reporting on the operation that we now refer to by Merlin’s codename.

And yet, I’m fairly certain, the fact that Merlin offered up Risen reporting to the man now convicted of having leaked to Risen, Jeffrey Sterling, 4 years before that leak began, never got mentioned at the trial.

Witness after witness in the Jeffrey Sterling trial made claims about how closely held the program was. “More closely held than any other program,” Walter C, a physicist who worked on the program described. “More closely held,” David Shedd, currently head of the Defense Intelligence Agency and head of Counterproliferation Operations until just after the Merlin op, said.

Of course, Bob S’ admission that — when FBI showed him a list, in 2003, of 90 people cleared into the program, he said it was incomplete — suggests all those claims are overstated.

But the details of just how careless the CIA was with Merlin’s identity raise further questions about claims that the operation — and especially Merlin’s identity — was closely held. Most striking to me is the revelation (Exhibit 17) that for months and months, Merlin was pitching his nuclear experience to Iran on an AOL account shared with his wife and kids. In the cable describing a January 12, 1999 meeting with Merlin (what appears to be the first where the two met alone) Sterling explained that Merlin had just opened a separate Hotmail account to use for his CIA spying.

“[M] also opened another email account through Hotmail. [M] opened the new account so his family, who also utilizes his AOL account, cannot access his email related to the project.”

That means from at least April 15, 1998 (see Exhibit 8, though Exhibit 16 suggests the effort started in November 1997), when Merlin started trying to make contact with Iranians who might be interested in a Russian nuclear scientist, until January 1999, Merlin’s contacts with Iran were completely accessible to his wife (who, given the evidence — as opposed to the sworn claims — presented in the trial, almost certainly knew anyway) and his kids (who may not have).

The AOL to Hotmail account switch appears to have been Merlin’s idea, but Sterling’s performance review for this period (Exhibit 60; note, it uses the name Samuel Crawford to protect Sterling’s identity) seems to reflect Sterling’s effort to train Merlin out of bad security habits. It says Sterling, “maintained a strong [counterintelligence]/security posture in all that he did during the reporting period, particularly a high interest sensitive case … and is constantly seeking to improve the security of his cases.” Indeed, the same cable covering the January 12, 1999 meeting in which Merlin and Sterling alone took part — which revealed Merlin had finally gotten a dedicated email account for his CIA work — also described Sterling walking Merlin through changes to his handling approach, including apparently meeting in hotel rooms rather than primarily restaurants all in the same neighborhood. Sterling also appears to have had to prompt Merlin to share all his correspondence with the Iranians with him (though Merlin didn’t always do so).

The CIA, it appears, intended for Merlin’s real identity to be readily obvious to the Iranians (and, based on the presumption at the heart of this operation that the Iranians were working closely with Russia, also to the Russians). On several occasions, defense attorneys asked CIA witnesses if they were “dangling” Merlin, a term the witnesses clearly wanted to avoid repeating. Nevertheless, it is clear they intended to dangle him, barely hiding his identity or location. Merlin used his real name in his outreach to Iran. He cited “his true background” in messages to Iranians (Exhibit 16; note that in Exhibit 17 and 18, CIA has actually redacted some of the information on himself Merlin sent to Iranians). He used his home email address in classified ads (and changed it when he got the Hotmail account). His approaches used a PO Box that appears to have been close to his home. (Exhibit 16)

Then there were other issues that raised alarms for me. From roughly October 12, 1998 through December 10, 1998, spanning the period when Merlin went to San Francisco for his “training” on the blueprints, Merlin’s home computer was being repaired (he accessed some email from work). When he got the computer back “it seemed slower.” Shortly after Merlin got the new email (Exhibit 18), he told Sterling he had been blocked out with an “Intrusion detected,” warning, and had been told “evidently at least two people had tried to open the account.” In February 1999 (Exhibit 21), Merlin told Sterling he was having problems with his AOL account (though didn’t explain what sort). Then there’s the period in October and November 1999 when Sterling couldn’t contact Merlin; he said he was visiting his wife in Florida. Admittedly, Sterling was tracking this stuff closely (I wonder whether the government had Merlin on what amounted to a consensual wiretap). Because the government only released cables from the period when Sterling handled Merlin, however, we can’t know whether the earlier, sloppier operational security extended to Merlin’s online life.

And ultimately, this loose operational security — and Merlin’s backlash to it — appears to be one of the things that led Merlin to botch the operation, to (apparently) give Iran what was meant to look like a sales proposal with no way for Iran to contact him.  A full year before the delivery (Exhibit 18) Merlin started using initials rather than his name in correspondence with the Iranians. He balked at the CIA’s insistence that he send a (slightly doctored) resume along with his outreach attempts. He started sending related letters via separate envelopes, even sending them from separate states (CIA doesn’t appear to want to hide that Merlin was in a position to send letters from both New Jersey and Connecticut).

It was very clear that Merlin did not want his identity associated with the documents in question and — because CIA had decided to have him finalize the letter on his own in Vienna — he chose not to leave much of it with the package. He left off his PO box from the package, which was how the Iranians were supposed to respond, which was one key to any ongoing intelligence gathering aspect to the operation. The CIA had Merlin leave a sales pitch with no way for interested buyers to act on that sales pitch.

And yet they called this operation a big success.

As the Senate Intelligence Committee’s report (Exhibit 101) on Jeffrey Sterling’s whistleblower complaint describes, “In the end, the entire plan was turned over to the Iranians without any means for further follow up. However, CIA supposedly deemed the operation a success.”  The government repeatedly claimed Sterling lied or spun facts to get the SSCI and Jim Risen interested in this story. But on this point — that Merlin left a nuclear blueprint wrapped in newspaper without an address to follow-up with — Sterling was absolutely correct.

As I suggested in my response to Walter Pincus’ laundering of false claims about the evidence presented at the Sterling trial, DOJ went to great lengths to help CIA try to rebut claims made in James Risen’s book about the blueprints handed over to the Iranians.

Most importantly, by confusing the issue of “incomplete” plans with “flawed” plans, they suggested no one — not Merlin and not Jeffrey Sterling and not anyone else — should have a concern about the nuclear blueprints Merlin wrapped in a newspaper and dropped in a mailbox. They mean something very limited by that: that no one identified the flaws CIA purposely put into the plans to ensure they wouldn’t work.

There are several concerns with that:

• Merlin’s concerns, both nuclear and operational
• Human Asset 2’s concerns
• Problems with their logic

I’m going to bracket Merlin’s concerns for the moment (DOJ seems unenthused about answering my questions on where I can get a transcript of Merlin’s testimony).

But there are plenty of reasons to be concerned for other reasons, too.

HA2’s minor details

First, there are potential concerns raised by the other Russian nuclear engineer, referred to as Human Asset 2, who provided the plan for a 1980s-era Russian fire set to the US in the first place. Scientists at a national lab reverse engineered it, then added in a bunch of flaws that would purportedly prevent the Iranians from being able to use it.

Walter C, who worked at the national lab, described working with the Russian and claimed that — in a series of four meetings at which he apparently did not have much time with the blueprints — he found nothing wrong with them. “It looked credible to him,” Walter C claimed on the stand.

Except Human Asset 2 does appear to have had some concerns. In a cable dated November 25, 1998, Bob S passed on information to the New York and one other office passing on what appears to be Walter C’s message that the two parts Merlin had identified to be missing from the fireset were not supposed to be included.

As we had suspected, the inclusion of certain assemblies on the parts list but not on the schematic was indeed intentional, with the goal of suggesting that the anonymous fireset designed knew that these two assemblies … were essential, but did not know how to make or spec them in any detail.

[snip]

Please advise [M] of this outcome and suggest that he plan to acknowledge this omission in his eventual presentation to the Iranians if they ask about it.

But the cable went on to reveal that there were other “minor details” that Merlin and HA2 — the original Russian designer of the fireset plans — had raised.

[Mr. G] is still authorized to travel to [the lab] to meet with members of the fireset team to look into some other minor details of the plans which M and [HA2] have noticed or questioned. But per agreement between CP officers and [Mr. C] we will make no changes to the plans and lists until a serious discrepancy arises.

In other words, HA2 may have said, on quick review, that the plans looked credible (at this point, Merlin had examined the plans for what appears to have been an even shorter period of time). But he did, at least, have questions, which is not something Walter C offered up in his testimony.

The Red Team’s 5-month 3-month nuclear project

Then there are other problems in their claims the blueprints wouldn’t be usable. In his testimony, Walter C described how a Red Team of the lab’s nuclear scientists — with 200 years of combined experience! Walter C boasted, as if that were a meaningful stat — tried to find flaws and use the blueprints themselves. While the team only found 25% of the flaws, Walter C claimed, they were able to get it to work in what in testimony (at least according to my notes) was just 5 months.

Only that’s not what the cable said. It said,

After three months of intense effort, and by finding and fixing [some of the] design flaws, the team was able to get a breadboard version of the fire set to work in a laboratory setting.

That’s not the same as using it for a nuclear weapon (as the cable goes onto explain), but at least by my hearing, Walter C misrepresented how long it took the lab to get these flawed blueprints to work.

The mixed assumptions about Russian involvement

Finally, there’s the other way witnesses — especially Walter C — dismissed concerns that the Iranians might be able to use the blueprints. The entire premise for using a Russian blueprint (albeit integrating American and Japanese parts) to distract the Iranians was the Russian brain drain: after the collapse of the Soviet Union, Iran started recruiting newly unpaid scientists. “They had websites looking for areas of WMD concern,” Bob S testified.

Thus, as they tried to build a fireset that the Iranians couldn’t use, they factored in that and “assumed assistance of Russian scientists.”

The export control approval (Exhibit 26) caveated that “in the very remote possibility that the end user can acquire the critical specifications intentionally omitted from the design, and if the user can also acquire the necessary fabrication technologies to successfully fabricate a fully functional device, the end product would then be subject to the above controls.” If the Iranians had access to Russian engineers, as assumed, how could the US be sure they couldn’t use this?

Meanwhile, the enabling technology letter (Exhibit 28) assumes the Iranians had more expertise than claims about their 1950s era nuke program may have let on. The enabling technology approval letter assumed that “User already has a basic understanding of either commercial or nuclear fireset design.”

We don’t have enough data to assess either of these claims (though Walter C’s credibility should be taken about as seriously as his claim that he is “only vaguely” aware of the 2007 NIE finding that Iran had no nuclear weapons program as of 2003). But there seems to be a logical problem with the claims surrounding the plan, that assume both that Iran has access to Russian know-how but that it also doesn’t have access to such know-how.

As I’ve said, I’m working on a larger post about what a shitshow the evidence entered into the Jeffrey Sterling trial showed the Merlin operation to be.

But before I do that, I want to look closely at how David Shedd’s sworn testimony (which according to him, he practiced with prosecutor Jim Trump three times before appearing) contradicts a detail in one of CIA’s cables, because I think it goes to the crux of CIA’s efforts to spin this as a successful program.

Before I do that, let’s review his background. From 1997 (just as the Merlin op started) until March 2000 (literally when the part that shows up in Risen’s book ends), Shedd was the Chief of Operations in the Counterproliferation Division. From March 2000 until February 20, 2001, he was CIA’s head of Congressional Affairs. Then, for over four years, he worked on proliferation issues at the National Security Council. On the stand, Shedd claimed that the NSC provided “real oversight” of intelligence. From 2005 until 2010, Shedd worked in the Office of Director of National Intelligence, ultimately as Deputy. Then he moved to the Defense Intelligence Agency, where he’s now the Acting Director.

In other words, Shedd had a supervisory role over the Merlin program until Merlin handed over the blueprints. Then, after a stint working with Congress, he helped Condi invent her mushroom clouds and was one of the people at NSC cleared into the program when, in 2003, Dr. Rice convinced NYT to kill the first Risen story.

In spite of his potential conflicts, Shedd ended up being the guy who provided a leak assessment for Chapter 9 of Risen’s book for Director of National Intelligence John Negroponte in 2006 (see page 11). Curiously, two parts of that leak assessment are redacted, meaning they have nothing to do with the Iranian op (though it could relate to the other countries CIA used Merlin to deal blueprints to, or to the exposure of all CIA’s Iranian sources described in the chapter).

The prosecution specifically asked Shedd how he was kept informed about the Merlin project. He said Bob S kept him informed in conversations, providing updates “as often as necessary,” and that he, Shedd, might see cable traffic. He also described relying on the National Lab’s assurances that the blueprints Merlin was handing over to Iran could not help their program. “As a non-specialist myself, I had to rely on those with a nuclear specialization.” (Almost all the CIA witnesses involved in Merlin said something similar, and they had really bizarre views on engineering expertise, which might be one reason the program ended up being a clusterfuck.)

Having laid out that background — particularly the bit about briefing in “conversations” with Bob S (who himself testified the CIA writes everything down) — I find a series of Shedd’s responses to Sterling attorney Edward MacMahon particularly interesting.  “You know the nuclear blueprints were delivered in a newspaper?” MacMahon asked. “I don’t know,” Shedd, one time advisor to both Condi and Negroponte responded. “You don’t know any details about how the blueprints were delivered?” MacMahon persisted. Merlin “had established contact through letters, he then had a meeting with the person in Vienna,” Shedd responded.

Of course, that’s wrong.

Even according to Bob S’ favorable description of the program, Merlin didn’t meet with anyone in Vienna. He just wrapped nuclear blueprints inside a newspaper with both a computer-written and a handwritten note of quasi-explanation and left them in a mailbox, apparently having taken his PO Box for further contact off the letter. So why was David Shedd — who had a supervisory role over this operation (what Bob S himself called one of the “Generals” who played a key “check and balance” over the program) and then went on to brief Condi and Negroponte on it — misinformed about such a critical detail of the case?

I find Shedd’s statement particularly interesting given that he is named in one of the two cables submitted into evidence on the outcome of the operation.

On March 10, 2000, Bob S wrote a cable (Exhibit 44; he claims Sterling may have been sitting at the next terminal while he wrote it), to Langley and CIA offices 5, 7, and 8.

Having finally located the mission after several very obvious searches in the vicinity, [M] at one point noted that there was someone in the office, but on that occasion he had not brought the document package with him. When he returned on two subsequent days he found the office unoccupied and finally left the package, very clearly addressed to [Iranian subject 1], in the locked mailbox right outside the mission door.

Much of the rest of the cable described what a hash Merlin had made of his delivery in Vienna, even describing Merlin’s “inability to follow even the simplest and most explicit instruction” (Bob S did leave some damning details out, and that assessment did not prevent Bob S from proposing the use of Merlin to do similar operations with other countries within a month).

Then, on March 13, 2000, Bob S wrote another cable (Exhibit 3) to CIA offices 9, 7, and 8, New York, and office 5 for information. Like the previous cable, it was titled “Mission Accomplished.” It asked those offices for any sign of an Iranian response to the blueprints. While the cable didn’t provide as much detail about what a bumbler Merlin was, it did explain,

Our asset visited the Iranian mission facility several times, but did not find any one present in the office on 2 or 3 March during his visits. He accordingly placed the packet in the locked mail box immediately adjacent to the door of the mission inside its host building at 19 Heinestrasse in Vienna’s Second District.

In this paragraph and in others, there’s significant spin. In this paragraph, for example, Bob S doesn’t reveal that Merlin did show up one day to find someone in the office, but claimed at the time he didn’t have the packet with him. But it does reveal a detail Shedd says he doesn’t know: that Merlin never actually met with any Iranian.

Now, it’s a tearline document, meaning the people in each of those offices are supposed to direct the information below the tear line in the cable to specific recipients within the office. The only thing most readers would see is an above line summary, one line of which is redacted here. But this is a document David Shedd signed off on the release of:

Of particular note, given that Shedd may have briefed other superiors about the program, when George Tenet talked James Risen out of reporting the story in 2003 (page 10), he said, “the Russian involved introduced himself to the Iranians [two words redacted].” Did Tenet tell Risen he introduced himself in person, as Shedd claims to have believed?

Now to be fair, one more cable Shedd signed off on (Exhibit 16) described a plan, dating to 1998, to have Merlin meet directly with the Iranians. So it’s possible Shedd simply remembers the operation as it was supposed to be, and not as Merlin’s bumbling execution carried it off.

I’m not sure what to make of the later cable, though. Perhaps Shedd never read beyond the tearline — though that would raise real questions about his level of knowledge of the operation and the “General’s” oversight over it generally. Perhaps he was on his way out of CPD and didn’t read that closely. Or maybe he didn’t sign off on the release of it at all.

But it does seem to suggest that, before Shedd left CPD, he was involved in a cable that made it clear that Merlin just left the nuclear blueprints in a mailbox. And yet the story Shedd now tells, and perhaps told Condi and Negroponte, is the story of the operation as it was meant to be, not as it was actually conducted.

Walter Pincus had a piece yesterday purporting to lay out the inaccuracies in the chapter of James Risen’s State of War. In it, he includes this passage.

In Vienna in late February 2000 to deliver the materials to an Iranian mission to the International Atomic Energy Agency, the Russian, according to Risen’s book, “unsealed the envelope with the nuclear blueprints and included a personal letter of his own to the Iranians. No matter what the CIA told him, he was going to hedge his bets. There was obviously something wrong with these blueprints — so he decided to mention that fact to the Iranians in his letter.”

Risen’s book reprints the letter, saying the Russian later gave the CIA a copy.

The CIA trial witnesses and agency memos tell a different story.

The agency plan always was that the schematics and drawings would have some obvious flaws — and the Russian engineer was told about them. It also was part of the plan from the start that the design materials were to be accompanied by a letter from the Russian noting some errors. A Jan. 10, 2000, CIA memo carries a draft of what it describes as “the letter to be included in the package of material.”

It has elements almost word for word found in the letter as printed in the Risen book, but it was written cooperatively with CIA input and made part of the document package for the Iranians more than a month before the Russian arrived in Vienna.

Now, I think the trial did show that there were some inaccuracies in the book — the one the Merlins cared most about is that they weren’t defectors.

But I find it really curious that Pincus claims these were errors. I say it’s curious because unless I’m mistaken, the transcripts for all the CIA witnesses save Bill Harlow have not been loaded onto the docket and so probably aren’t yet done. And in the 5 of 6 days of testimony I attended (including all but a few minutes of Bob S’ testimony, whom Pincus cites by name), I didn’t see Pincus in the courtroom once. And with the exception of Merlin himself, the CIA witnesses I missed, for the most part, talked about issues other than the Merlin operation. So it’s unclear where Pincus got his understanding of CIA witness testimony, and what he got is inaccurate.

Indeed, in this limited example, Pincus makes two pretty significant errors: in suggesting Merlin was supposed to know about the flaws in (as opposed to the incompleteness of) the blueprints, and in suggesting the CIA is certain about what Merlin left at the IAEA in March 2000.

First, the flaws. Throughout discussions about this operation, there has been some confusion between the flaws and the incompleteness, which has allowed the CIA to push back on the story when in fact the CIA records show this may be a convenient way to claim Risen’s book was wrong when what the CIA thought is meaningless if the Russians still had concerns. While Merlin was told the blueprints were incomplete, he was not told about the flaws the nuclear lab (probably Sandia) put in the blueprints that were supposed to prevent the Iranians from using them (but only held back a national lab team 3 months in using the same blueprints). According to my notes, for example, Bob S said they “didn’t want to say [the blueprints] were intentionally flawed,” to Merlin. Nevertheless, there is reason to believe that Merlin and (far more importantly) the other Russian asset involved in this operation saw what they believed were problems that would make the blueprints not serve the purpose the Russians believed they were supposed to serve, and there is reason to believe that those concerns were never adequately addressed.

In addition, as I noted in this Salon piece yesterday, CIA doesn’t actually have the final version of what Merlin left with the IAEA. They claim — with questionable credibility, which I’ll return to — not to know what was in the formal letter Merlin left. Bob S himself agreed in his testimony that Pincus supposedly reviewed that Merlin is the only person who knows what he put in the final version. At the very least the story the CIA tells is that Merlin took a copy of the letter drafted in conjunction with the CIA to Vienna but with the nuke references altered to make sure he could get through customs (Bob S called it “sanitized”), then changed them back on the hotel computer and printed a fresh copy (note, earlier in this process, Merlin at times sent stuff off to the Iranians before the CIA had a chance to review it, so he had a history of freelancing). He then destroyed the disk he used, meaning no one — according to what Merlin told CIA  — has a copy (though the almost-final version without any last minute changes would reside on Merlin’s poorly secured home computer). Interestingly, Risen’s book says Merlin wrote a report back, but Bob S and Merlin (apparently) claim he did not.

But that printed letter is not all Merlin left with the blueprints. He also left a handwritten letter in his  packet of newspaper-wrapped nuclear blueprints — what Bob S called a “cover note.” The current story — relying on an earlier idea floated during the drafting period but not formally adopted — is that the cover note would help alert the Iranian staffers to the ultimate intended recipient of the letter. But that letter was by all appearances ad-libbed by Merlin. So we only have Merlin’s word for what he wrote.

Now these are just two details — details in Risen’s book that Pincus claims were disproven by cables and Bob S’ testimony — but which were anything but.

I will have a much longer summary of all the other details that came out at trial that made it clear the operation was an even bigger shitshow than Risen’s report makes out. But for the moment, I’m just curious what Pincus is trying to accomplish. Perhaps he was in the back of the courtroom for a tiny part of Bob S’ testimony and neither I nor the several other journalists I asked noticed him. But (at least as far as testimony) it appears he’s working off second-hand claims about what the record says and claiming, falsely, that they specifically disprove Risen’s book.

Why?

Why would whoever provided Pincus this partial view of Bob S’ testimony be so desperate to claim that Risen’s book was proven wrong?

In a piece for Salon, I describe how the government managed to get Jeffrey Sterling convicted of 7 charges under the Espionage Act for one leak. More importantly, I show how the jury’s conviction of him for 2 of those charges — related to “causing” James Risen to write a 2003 NYT story on Merlin that got quashed — may well amount to convicting him for tipping Risen, without sharing any classified information, to the operation.

Here’s the key part of that discussion:

D.C. information brokers should be worried that Sterling faces 80 years in prison based off this circumstantial evidence. All the more so, given the evidence supporting the charge that Sterling leaked to Risen in time for and caused him to write the article Risen told CIA he had in completed draft on April 24, 2003. After all, the only pieces of evidence that the government submitted from before the time when Risen told CIA he had a completed article were the CNN email, phone calls reflecting Risen and Sterling spoke for four minutes and 11 seconds across seven phone calls, and Sterling’s entirely legal discussion with staffers from the Senate Intelligence Committee.

No matter what you think all the later phone calls between Sterling and Risen indicate, short of evidence of a face-to-face meeting in this earlier period, the evidence seems to suggest Sterling was doing something that people in DC do all the time: point an investigative reporter to where she might find classified scoops, without providing those scoops themselves. That’s especially true given the way the CIA’s own notations about Risen’s story seem to track the reporter fleshing out information, from initial outlines of the operation (that happen to map what Sterling told Senate staffers) to, weeks later, inclusion of that elusive document FBI never managed to find. That is, it appears Risen got a tip, possibly from Jeffrey Sterling, but that he spent weeks using his sources to flesh out that tip.

In both the indictment and discussions about jury instructions, the government interpreted the Espionage Act to cover what might be an unclassified tip through two means. First, they pointed to language in the Espionage Act that criminalizes someone “caus[ing secrets] to be communicated, delivered, or transmitted,” and from that argued Sterling was responsible not just for the leak to Risen but also for the journalist’s attempt to publish a newspaper article and his completion and his publisher’s delivery to Virginia of a book chapter. Then, for most counts, they argued that Sterling did not have to have handed Risen secret information directly, he could do so indirectly.

If the jury found Sterling indirectly got secrets into Risen’s hands and, from that, caused him to write an article and a book chapter on it (irrespective of the additional work Risen did, the work of his editors at the Times and the publishers at Simon and Schuster and the commercial freight company that carried those secrets in a bound book to Virginia), that was enough to send him to prison for most of the rest of his life.

While it’s all well and good that DOJ backed off plans to force James Risen to testify, I think few realize the implications of Sterling being held responsible for an entire NYT story based on four minutes and 11 seconds of phone conversations.

They may well criminalize providing unclassified tips to get reporters to chase down classified stories.

Yesterday, WikiLeaks revealed it had just been notified by Google that DOJ subpoenaed all the records of 3 staffers 2 years ago. Today, WaPo provides some details on Google’s efforts to fight the gag orders that have prevented them from informing customers about such subpoenas.

It reveals that EDVA imposed a gag for WikiLeaks subpoenas after the Twitter one that caused such a stink years ago because “the resulting publicity” from that Twitter subpoena “was a disaster for them.”

“The U.S. attorney’s office thought the notice and the resulting publicity was a disaster for them,” Gidari said. “They were very upset” about the prosecutor’s name and phone number being disclosed, he said. “They went through the roof.”

About the same time that the Twitter story broke, Google was served with a separate order for the data of Jacob Appelbaum, a WikiLeaks volunteer and security researcher. Google wanted to inform him, but prosecutors balked.

“There was a lot of pushback from the government because they also were getting pressure from the people who got served from Twitter,” Gidari said. “The U.S. attorney’s office is like, ‘Hell no, we’ll fight you forever.’ ”

For the next four years, “Google litigated up and down through the courts trying to get the orders modified so that notice could be given,” he said.

As EFF’s Hanni Fakhoury noted on Twitter, this basically means the government is arguing that some harassment and popular criticism about the decision to subpoena WikiLeaks amounts to an adverse result:

(2) An adverse result for the purposes of paragraph (1) of this subsection is—
(A) endangering the life or physical safety of an individual;

(B) flight from prosecution;

(C) destruction of or tampering with evidence;

(D) intimidation of potential witnesses; or

(E) otherwise seriously jeopardizing an investigation or unduly delaying a trial.

I’m not excusing harassment, but DOJ is effectively prioritizing avoiding criticism over rights to notice, including (especially in the case of Jacob Appelbaum) to an American citizen.