There's a legal question in U.S. v. Bourke—a case in the 2nd Circuit—that's fascinating (and not just because it bridges my two worlds: anti-corruption and discovery). Like a lot of legal arguments in real-life cases, you have to peel back layers of legal detritus and lawyer bickering to get to it. And as in any real-life case, the facts are sometimes clear, sometimes murky and generally convoluted.

Let's start with what we know for sure: Frederick Bourke was convicted after trial of participating in a scheme to bribe Azeri government officials relating to the possible privatization of Azerbaijan's oil industry.

A major issue in the case is how much Bourke knew about the bribery. Bourke claimed he was just an innocent investor, taken in by the “real” bad guys. Even Judge Shira Scheindlin said that “after years of supervising this case, it's still not entirely clear to me whether Mr. Bourke is a victim or a crook or a little bit of both.” Scheindlin sentenced Bourke to a year and a day in prison.

At trial, the government presented testimony from a Swiss lawyer, who established that Bourke knew about the bribery and knew about it before he invested his money. It is in that testimony that the seeds of our legal issue first took root.

The Swiss lawyer, Hans Bodmer, testified that on the afternoon of Feb. 5, 1998, he and Bourke were both in Baku (the capital city of Azerbaijan). Bourke, said Bodmer, approached him in the lobby of the Baku Hyatt and asked for an update on the investment. That night, Bodmer spoke with another co-conspirator and got permission to lay out the bribery aspects of the transaction to Bourke.

Early the next morning, Bodmer and Bourke went for a walk, and Bodmer told Bourke the whole story, including about the bribery scheme. This interaction was dubbed the “walk talk.” Approximately two weeks later, Bourke invested $7 million of his and other people's money in the investment.

Flight records, however, make Bodmer's account impossible. On Feb. 5, Bourke was in London, not Baku, and so he couldn't have been at the Baku Hyatt as Bodmer testified. Bourke didn't arrive in Baku until the morning of the Feb. 6, after Bodmer's “walk talk” was supposed to have happened.

Prosecutors were unaware of the contents of these flight records, despite having custody of them for several years. Apparently, either no one read them or no one understood the consequence of them.

If the government knew or should have known that its key witness was making up a major piece of the case, does the conviction need to be overturned? The law seems to indicate yes.

I don't think anyone is claiming the government actually knew about the contradiction. Had it known, it wouldn't have made such a big deal of the “walk talk” during its opening statements, or would have at least mentioned the witness's misremembering the date.

If the government should have known about Bodmer's alleged lie (so-called by the defense … the government says it was just a mistake of timing after several years had passed), the appellate court can still set aside the jury verdict.

But can we truly say that the government should have known the contents of the flight records?

This is the legal question I find fascinating. Is a party charged with knowledge of documents in its possession if it never looked at the documents? If so, then the government “knew” of the contradiction when it put Bodmer on the stand. The defense claimed that because the government had the document, it knowingly introduced false testimony. One uncharitable term for that is “suborning perjury.”

The flight records were turned over in a disclosure that even Scheindlin called “voluminous.” Robertson Park, a former Department of Justice (DOJ) attorney—now a Partner at Murphy & McGonigle—who served on the DOJ's prosecution team, said that the disclosures were “very substantial in scope, into the hundreds of thousands of documents, and it wouldn't stun me if it was in the higher end of that.”

In the modern discovery environment, in which document production volumes are often so large that it is literally a mathematical impossibility to review them all, how can we impute knowledge?

It doesn't help the government's case that the document was not only in its possession, but also was turned over to the defense. On the civil side, where discovery volumes can easily reach into the millions of documents, the act of production itself has implications. Bennett Borden, chair of the e-discovery and information governance section at Williams Mullen and vice chair of the American Bar Association's e-discovery and digital evidence committee, is one of the nation's leading experts on e-discovery, and he says that the production of documents “adds a layer of complexity.” “When a party produces a document,” he says, “they're asserting that it's responsive to a request. By turning it over, there's an assumption that they've at least looked at the document in order to make that assertion.”

Even in this environment, Bennett said, parties can agree to turn over documents en masse (after filtering out potentially privileged documents), without reviewing the remainder. These types of agreements make civil discovery much simpler and less expensive.

That simplicity and cost savings is true for the government as well. In the Bourke case, according to Park, “we turned over pretty much everything. It was too cumbersome to make relevance determinations, so we just turned over everything we had. If it was provided by a third party, we turned it over. It was too much to parse through. Having done a number of long-term investigation document productions, it's not always worth the time, energy and effort to make a more nuanced production of materials.”

This explains how it is that the government didn't know about a document in its possession for years, and which document it turned over to the defense. It's an eminently believable explanation, in my opinion. Case law seems to indicate that when alerted to problems with its witnesses, the government has a duty to investigate. But that's not the case here. No one knew about the problem with the flight records until after Bodmer testified. Reported decisions in this area are sparse.

Bennett explains that “the law is reactive by nature. With every significant technological advance, the law takes time to catch up. Before the information age, with its consequent data deluge, it might have been fair to assume a party knew what was in the documents it produced. I'm not sure that is the case now.” This is true for this legal issue as well. There is one case I've found that's almost on point, U.S. v. Cribbs, a September 2011 case out of the Western District of Tennessee.

In Tennessee, in 2009, a man called the Bureau of Alcohol, Tobacco, Firearms and Explosives and said his father was acting erratically and had firearms in the house. The police responded, and the son met them at the door. The son consented to a search of the house and opened the gun safe using a numeric code.

The problem is, the father had that same day gotten an order of protection against the son. The order required the son to stay away from the house. The defendant said that because the order was in the police's computer system, knowledge of it should be imputed to the officer who used the son's consent to search the house. And “knowing” about the order, the officer should not have searched the house without a warrant; the son's “consent” was invalid.

The magistrate judge found for the defendant. The district court judge overruled the magistrate judge.

Defendant asks the Court to impute knowledge of information contained in a law enforcement database to officers conducting an investigation in the field. Defendant has not provided any case law that supports such an application of the imputed knowledge doctrine, nor has the Court found any.

I would argue that as data volumes rise, the idea of imputing knowledge to a corporation, police force or the government generally makes less and less sense.

Even within corporations, contract attorneys often conduct document review. These workers aren't looking to develop the document record; they look through massive amounts of documents as quickly as possible for the limited purpose of seeing whether a document is minimally responsive to a document request. Just because a contract attorney sees a document shouldn't impute knowledge to the law firm that hired them, and then on to the client, in my opinion.

But, remarked Park, “I could imagine a time, given future advances in search technology, when parties may be held responsible for what's in their document collections.” Advanced concept searches and review processes such as predictive coding make that scenario more likely, but we're not there yet.

There's a legal question in U.S. v. Bourke—a case in the 2nd Circuit—that's fascinating (and not just because it bridges my two worlds: anti-corruption and discovery). Like a lot of legal arguments in real-life cases, you have to peel back layers of legal detritus and lawyer bickering to get to it. And as in any real-life case, the facts are sometimes clear, sometimes murky and generally convoluted.

Let's start with what we know for sure: Frederick Bourke was convicted after trial of participating in a scheme to bribe Azeri government officials relating to the possible privatization of Azerbaijan's oil industry.

A major issue in the case is how much Bourke knew about the bribery. Bourke claimed he was just an innocent investor, taken in by the “real” bad guys. Even Judge Shira Scheindlin said that “after years of supervising this case, it's still not entirely clear to me whether Mr. Bourke is a victim or a crook or a little bit of both.” Scheindlin sentenced Bourke to a year and a day in prison.

At trial, the government presented testimony from a Swiss lawyer, who established that Bourke knew about the bribery and knew about it before he invested his money. It is in that testimony that the seeds of our legal issue first took root.

The Swiss lawyer, Hans Bodmer, testified that on the afternoon of Feb. 5, 1998, he and Bourke were both in Baku (the capital city of Azerbaijan). Bourke, said Bodmer, approached him in the lobby of the Baku Hyatt and asked for an update on the investment. That night, Bodmer spoke with another co-conspirator and got permission to lay out the bribery aspects of the transaction to Bourke.

Early the next morning, Bodmer and Bourke went for a walk, and Bodmer told Bourke the whole story, including about the bribery scheme. This interaction was dubbed the “walk talk.” Approximately two weeks later, Bourke invested $7 million of his and other people's money in the investment.

Flight records, however, make Bodmer's account impossible. On Feb. 5, Bourke was in London, not Baku, and so he couldn't have been at the Baku Hyatt as Bodmer testified. Bourke didn't arrive in Baku until the morning of the Feb. 6, after Bodmer's “walk talk” was supposed to have happened.

Prosecutors were unaware of the contents of these flight records, despite having custody of them for several years. Apparently, either no one read them or no one understood the consequence of them.

If the government knew or should have known that its key witness was making up a major piece of the case, does the conviction need to be overturned? The law seems to indicate yes.

I don't think anyone is claiming the government actually knew about the contradiction. Had it known, it wouldn't have made such a big deal of the “walk talk” during its opening statements, or would have at least mentioned the witness's misremembering the date.

If the government should have known about Bodmer's alleged lie (so-called by the defense … the government says it was just a mistake of timing after several years had passed), the appellate court can still set aside the jury verdict.

But can we truly say that the government should have known the contents of the flight records?

This is the legal question I find fascinating. Is a party charged with knowledge of documents in its possession if it never looked at the documents? If so, then the government “knew” of the contradiction when it put Bodmer on the stand. The defense claimed that because the government had the document, it knowingly introduced false testimony. One uncharitable term for that is “suborning perjury.”

The flight records were turned over in a disclosure that even Scheindlin called “voluminous.” Robertson Park, a former Department of Justice (DOJ) attorney—now a Partner at Murphy & McGonigle—who served on the DOJ's prosecution team, said that the disclosures were “very substantial in scope, into the hundreds of thousands of documents, and it wouldn't stun me if it was in the higher end of that.”

In the modern discovery environment, in which document production volumes are often so large that it is literally a mathematical impossibility to review them all, how can we impute knowledge?

It doesn't help the government's case that the document was not only in its possession, but also was turned over to the defense. On the civil side, where discovery volumes can easily reach into the millions of documents, the act of production itself has implications. Bennett Borden, chair of the e-discovery and information governance section at Williams Mullen and vice chair of the American Bar Association's e-discovery and digital evidence committee, is one of the nation's leading experts on e-discovery, and he says that the production of documents “adds a layer of complexity.” “When a party produces a document,” he says, “they're asserting that it's responsive to a request. By turning it over, there's an assumption that they've at least looked at the document in order to make that assertion.”

Even in this environment, Bennett said, parties can agree to turn over documents en masse (after filtering out potentially privileged documents), without reviewing the remainder. These types of agreements make civil discovery much simpler and less expensive.

That simplicity and cost savings is true for the government as well. In the Bourke case, according to Park, “we turned over pretty much everything. It was too cumbersome to make relevance determinations, so we just turned over everything we had. If it was provided by a third party, we turned it over. It was too much to parse through. Having done a number of long-term investigation document productions, it's not always worth the time, energy and effort to make a more nuanced production of materials.”

This explains how it is that the government didn't know about a document in its possession for years, and which document it turned over to the defense. It's an eminently believable explanation, in my opinion. Case law seems to indicate that when alerted to problems with its witnesses, the government has a duty to investigate. But that's not the case here. No one knew about the problem with the flight records until after Bodmer testified. Reported decisions in this area are sparse.

Bennett explains that “the law is reactive by nature. With every significant technological advance, the law takes time to catch up. Before the information age, with its consequent data deluge, it might have been fair to assume a party knew what was in the documents it produced. I'm not sure that is the case now.” This is true for this legal issue as well. There is one case I've found that's almost on point, U.S. v. Cribbs, a September 2011 case out of the Western District of Tennessee.

In Tennessee, in 2009, a man called the Bureau of Alcohol, Tobacco, Firearms and Explosives and said his father was acting erratically and had firearms in the house. The police responded, and the son met them at the door. The son consented to a search of the house and opened the gun safe using a numeric code.

The problem is, the father had that same day gotten an order of protection against the son. The order required the son to stay away from the house. The defendant said that because the order was in the police's computer system, knowledge of it should be imputed to the officer who used the son's consent to search the house. And “knowing” about the order, the officer should not have searched the house without a warrant; the son's “consent” was invalid.

The magistrate judge found for the defendant. The district court judge overruled the magistrate judge.

Defendant asks the Court to impute knowledge of information contained in a law enforcement database to officers conducting an investigation in the field. Defendant has not provided any case law that supports such an application of the imputed knowledge doctrine, nor has the Court found any.

I would argue that as data volumes rise, the idea of imputing knowledge to a corporation, police force or the government generally makes less and less sense.

Even within corporations, contract attorneys often conduct document review. These workers aren't looking to develop the document record; they look through massive amounts of documents as quickly as possible for the limited purpose of seeing whether a document is minimally responsive to a document request. Just because a contract attorney sees a document shouldn't impute knowledge to the law firm that hired them, and then on to the client, in my opinion.

But, remarked Park, “I could imagine a time, given future advances in search technology, when parties may be held responsible for what's in their document collections.” Advanced concept searches and review processes such as predictive coding make that scenario more likely, but we're not there yet.