Discovery in the age of terabytes

The CEO makes a frantic call to her attorney at 10:32 a.m. She says federal agents have just arrived at the company's offices, search warrant in hand, demanding entry. A handful of agents head straight for the server room. They're telling the tech geeks to shut everything down, slapping evidence labels on server towers and desktops, and carting them out.

Fast-forward six months. The company is in a shambles. The CEO has been indicted for Medicare fraud, and the government has just delivered to defense counsel discovery, in the form of 24 hard-drives of data seized during that office search. Defense resources are modest, and counsel now has three terabytes of data to get through in the five months that stand between now and trial. That's more than 1.5 billion pages of text. You'd have to review 10 million pages a day to get through it by trial.

Now here's the question: Does the government's data deluge satisfy its discovery duties? Are the government's discovery obligations met when defense counsel is buried in bytes?

As a practical matter, the answer is clearly "no." But as a legal matter, the question is an open one. Federal courts have danced deftly around the issue, and federal prosecutors have generally done just enough to avoid an adverse ruling.

But the day is coming when the government and the courts will have to deal squarely with a question that was inconceivable in 1963 when the seminal criminal discovery case of Brady v. Maryland was decided: How can the government comply with its constitutionally based discovery obligations in an age of terabytes?

Simply dumping a digital haystack and a few needles on the defense can't be enough. Instead, criminal discovery practices must be guided by the principle of ensuring that the accused is provided with meaningful access to discovery.

Courts are all over the map on this front. Many give short shrift to defense complaints about data dumps. Some more sympathetic courts approach the issue from the standpoint of time and preparation - that is, they will require discovery to be produced sufficiently in advance of trial to allow the defense to review it. But as with the foregoing frantic CEO example, just offering more time is sometimes not enough because the amount of data is just too great to permit any sort of manual review. But most all courts stop short of demanding that the government take robust steps to afford the defense meaningful access to the mountains of digital information produced in discovery.

In the absence of pressure from the courts, the government reflexively champions miserly discovery practices, insisting that turning over raw data is enough under Brady and Rule 16, and resisting any request to do more.

Occasionally, in cases where the kilobyte count is sky-high, the government reluctantly offers to identify what it considers to be "key docs." But those are often just the documents the government thinks are inculpatory, and are of little help to the defense, which seeks exculpatory material the government didn't find or didn't look for. A "key docs" approach is no solution at all to the meaningful access problem.

However, in a handful of cases, the government has done the right thing - indeed, the only thing it can do to ensure meaningful access to voluminous electronic discovery: it has loaded all the data in its possession into a searchable database and given the defense access to it in discovery. This rarely implemented but critically important approach to discovery in high-volume discovery cases is an approach whose time has come. It is the virtual version of the gold standard for criminal discovery, known as "open file" discovery.

Prosecutors shouldn't resist this kind of discovery - they should embrace it. A prosecutor who thinks she has a righteous case should be unafraid to reveal her case to the defense, warts and all, and allow it to be subjected to a level of scrutiny and challenge commensurate with the burden of proof in every case, beyond a reasonable doubt. Such liberal discovery practices also have the virtue of quelling nasty discovery battles and avoiding appellate reversals.

Despite the wisdom of an open-file approach to digital discovery, the U.S. Department of Justice's recently issued guidance on electronic discovery practices in criminal cases doesn't even mention the discovery database option. Instead, it speaks in the traditional language of discrete document productions and pays lip service to coordination. It's a real disappointment that the DOJ would leave out of its data discovery guidance the discovery approach that promises the greatest access and fullest disclosure.

Imagine if, rather than resisting this kind of approach, the DOJ adopted a practice of making discovery available on a searchable, shared database in every document-intensive case. Defense counsel would have immediate access to critical discovery. Prosecutors would spend less time on burdensome, one-off document productions. Courts would be less burdened by discovery battles. And defendants' rights to discovery in criminal cases would be far more fully secured than they are now.

The day is coming when a court will require the government to undertake such discovery practices as a matter of course. The question is whether the government will wait for that day, or do the right thing before that day comes.