Category Archives: eDiscovery

PROPOSED STATE BANS ON PHONE ENCRYPTION MAKE ZERO SENSE

 

Lock_Case
American politics has
long accepted the strange notion that just a pair of states—namely Iowa and New Hampshire—get an outsize vote in choosing America’s next president. The idea of letting just two states choose whether we all get to have secure encryption on our smartphones, on the other hand, has no such track record. And it’s not a plan that seems to make much sense for anyone: phone manufacturers, consumers, or even the law enforcement officials it’s meant to empower.

Last week, a California state legislator introduced a bill that would ban the retail sale of smartphones with that full-disk encryption feature—a security measure designed to ensure that no one can decrypt and read your phone’s contents except you. The bill is the second piece of state-level legislation to propose that sort of smartphone crypto ban, following a similar New York state assembly proposal that was first floated last year and re-introduced earlier this month. Both bills are intended to ensure that law enforcement can access the phones of criminals or victims when their devices are seized as evidence.

If consumers will cross borders to fill a booze cabinet, what’s to prevent New York criminals from foiling surveillance with New Jersey iPhones?

Those two proposed crypto bans have put another twist in an already tangled debate: The privacy and cryptography community has long opposed any such “backdoor” scenario that gives cops access to encrypted smartphones at the risk of weakening every device’s data protections. But legal and technical experts argue that even if a national ban on fully encrypted smartphones were a reasonable privacy sacrifice for the sake of law enforcement, a state-level ban wouldn’t be. They say, the most likely result of any state banning the sale of encrypted smartphones would be to make the devices of law-abiding residents’ more vulnerable, while still letting criminals obtain an encrypted phone with a quick trip across the state border or even a trivial software update.

Crypto Has No Borders

If the New York and California smartphone encryption bans passed, a company like Apple that sells encrypted-by-defaulted iPhones would have three options, argues Neema Singh Guliani, an attorney with the American Civil Liberties Union: It could cease to fully encrypt any of its phones, contradicting a year of outspoken statements on privacy by its CEO Tim Cook.  It could stop selling phones in two of America’s richest states. Or finally, it could create special versions of its phones for those states to abide by their anti-encryption laws.

The last of those scenarios is Apple’s most likely move, says Singh Guliani, and yet would result in a “logistical nightmare” that still wouldn’t keep criminals from encrypting their phones’ secrets. She compares the laws to state-wide liquor regulations: “People will travel over the border to buy alcohol in states with the standards that suit them,” she says. If consumers will cross borders to fill a booze cabinet, what’s to prevent New York criminals from foiling surveillance with New Jersey iPhones? “Nothing would stop those who wanted a more privacy protective phone to get one from out of state.”

In the hypothetical future where the state bills have passed, fully encrypting an iPhone might not even require buying an out-of-state device, but merely downloading out-of-state firmware. After all, it’s unlikely Apple would go to the expense of manufacturing different hardware for its phones to disable encryption in some of them, argues Jonathan Zdziarski, an iOS forensics expert who has worked with police to decrypt phones. “That would be a massive technical change to support this kind of device,” Zdziarski argues. “It would be literally cheaper for Apple to stop selling phones in California altogether.” Instead, he says, it would likely sell the same hardware for all of its devices and merely disable full-disk encryption through a different version of its firmware activated at the time of the phone’s purchase. And nothing in the current bills would prevent Apple from making the fully encryption-enabled version of its firmware available to anyone who restores their device from factory settings.

The technologically savvy will find ways to get encryption, while the average smartphone user’s data will be left more vulnerable.

In other words, that would make the New York and California crypto bans statewide bans on software, an idea roughly as practical as policing undocumented birds crossing the Mexican border. And if Apple were to try to accommodate the spirit of the law by preventing customers from restoring their phone with full-disk encryption inside California or New York, Zdziarski is confident iPhone owners could circumvent any location tracking, proxying their IP address or putting the phone in a Faraday bag to block its GPS. “This legislation is going to be technologically useless,” says Zdziarski. “Anyone who wants a device that doesn’t have law-enforcement-reversible encryption will be able to get one.”

Pressuring Congress

Neither Apple nor Google, which followed Apple’s lead last year by declaring that all devices running the latest version of Android will have default full-disk encryption, responded to WIRED’s request for comment on the California or New York bills. The office of New York Assemblyman Matthew Titone, who introduced the New York bill, tells WIRED that the state-level bill is meant to pressure Congress to follow with its own legislation. “When there’s no national legislation, states take efforts on their own to solve an issue,” says Titone’s chief of staff Chris Bauer. “That can speed the process along to make the federal government take steps.”

Skyler Wonnacott, the director of communications for the California bill’s sponsor Assemblyman Jim Cooper, offered a similar argument. “California is leading the fight…It’s got to start somewhere,” Wonnacott says. “Just because you can drive into Nevada and buy a phone or download software doesn’t mean there isn’t an issue and these phones aren’t used in crimes.”

Congress has yet to introduce legislation to limit full-disk encryption in smartphones, despite several congressional hearings over the last year in which officials, including FBI Director James Comey and New York District Attorney Cyrus Vance, warned of the dangers of allowing criminals access to devices with data they couldn’t decrypt. (Vance said at the time that New York police had been stymied by smartphone encryption 74 times in the nine months before the hearing, out of roughly 100,000 cases it deals with in a year.) A spokesperson in Vance’s office writes to WIRED that the DA’s office pushed for state legislation, and still hopes to find a compromise with device makers. “When Apple and Google announced the switch to full-disk encryption…with no regard for the effect it would have on local law enforcement and domestic crime victims, they left us with no choice but to seek legislative solutions at all levels, state and federal,” writes the district attorney’s director of communications Joan Vollero. “If the companies have a solution, we encourage them to engage in a productive dialogue.”

Constitutional Questions

But even if state laws do put pressure on Apple and Google to cave on encryption, they may do so unconstitutionally, says Andrew Crocker, an attorney with the Electronic Frontier Foundation. He says statewide smartphone encryption bans may fall under the “dormant Commerce Clause,” which gives the exclusive right to regulate commerce between states to the federal government. “States don’t have unlimited power to enact regulations to burden interstate commerce,” says Crocker. “If I’m Apple, this seems like a huge burden on my business.”

Congress, on the other hand, would have the power to ban default full-disk encryption in smartphones—though they’d do so against the advice of nearly every technical expert in the field of cryptography. In July of last year, for instance, 15 renowned cryptographers published a paper cautioning against any deliberate weakening of encryption for the sake of law enforcement. “New law enforcement requirements are likely to introduce unanticipated, hard to detect security flaws,” the paper reads. “The prospect of globally deployed exceptional access systems raises difficult problems about how such an environment would be governed and how to ensure that such systems would respect human rights and the rule of law.”

And Crocker reiterates that state-level bills wouldn’t be just problematic or risky, but “wildly ineffective,” as those who want encryption will easily get it from out of state—in either software or hardware form. The technologically savvy will use it to defeat police surveillance or to protect their phone from hackers and thieves, while the average smartphone user’s data will be left more vulnerable. “The ones who will actually be impacted are the less sophisticated people who don’t know how to get this protection,” says Crocker. “You’re looking at a cost that falls on innocent people, not criminals or terrorists.”

Judge David Campbell: Predictive coding rarely proposed, rarely used

Capture

This is the second part of an interview with US District Court Judge David Campbell, the former chair of the advisory committee responsible for drafting newly promulgated amendments to the Federal Rules of Civil Procedure. In part one, Judge Campbell addressed the prospective impact of the rule changes, and what their emphasis on proportionality and cooperation may mean in practice. He also outlined the evolution of the federal spoliation sanction rule, 37(e), which has been the focus of much debate and handwringing.

Here he discusses what other measures must accompany the rule changes to bring a substantive reduction in litigation costs, speed case resolutions, and reopen the federal court system to those it has priced out. We also ask him to share his experience with predictive coding and how it is — or is not — being used in his courtroom.

Logikcull: In the commentary to Rule 37, the committee noted that “The court should be sensitive to the party’s sophistication with regard to litigation in evaluating preservation efforts; some litigants, particularly individual litigants, may be less familiar with preservation obligations than others who have considerable experience in litigation.” Has anybody challenged that language as potentially giving a free-pass to litigants who are “willfully ignorant” — or maybe just lazy — when it comes to their preservation obligations?

Hon. David Campbell: I have not heard those challenges. It may well be that some folks have that concern. I will tell you, from my perspective, a very important background for this rule change is that the reality is that ESI in now in the possession of everybody. The personal injury plaintiff who walks into the lawyer’s office on crutches after the car accident has ESI that’s relevant to the injury, whether it’s their Facebook page, or the text they sent to their girlfriend after the accident, or communications with their doctors, or emails they might have sent. And that’s a very different world than we lived in 20 years ago. The reason for the comment you just read from the committee note is that if that person turns out to have not stopped Facebook’s deletion of posts — and I don’t know how Facebook deletes posts, but just use that as a hypothetical — a court can take into account their lack of sophistication in deciding later what Rule 37(e) measures should be imposed.

And I think that’s right. I don’t think we should hold them to the same standard as an entity that has an IT department. A number of the most prominent cases on the loss of ESI — I believe this is correct — deal with the plaintiff’s loss of ESI.

We had an expert tell us in one of these (rules committee) hearings that by 2018 — maybe he said 2020 — there will be 26 billion devices on the internet, which is, you know, four for every person on Earth. And the truth is, I believe, five years from now, 10 years from now, the amount of information that each person has in the cloud will be equivalent to the kinds of records that use to be found in the filing cabinets of entire businesses. So ESI, in my view, isn’t just a problem for the big entities. It’s a litigation issue for everyone, and this rule (Rule 37(e)) tries to take that into account.

“Five years from now, the amount of information that each person has in the cloud will be equivalent to the kinds of records that use to be found in the filing cabinets of entire businesses. So ESI, in my view, isn’t just a problem for the big entities. It’s an issue for everyone.”

Logikcull: Well, to that point, data growth is accelerating at an incredible rate. Is there any fear that whatever cost-lowering impact these changes will ultimately have will be negated by the fact that, not only is there going to be exponentially more data, but also that the technology available to handle that data does not appear to be getting much cheaper? Do you have that concern? I imagine you do.

Judge Campbell: I do. I think any judge or any lawyer involved in litigation should. We are hoping that these rule changes help through Rule 37(e) in bringing some level of uniformity of how you deal with the loss of information. We hope Rule 37(e) will reduce the amount of side litigation that occurs over loss of ESI and sanctions. The proportionality change and the case management changes, as well, are intended to get judges involved earlier in cases and work with the parties in figuring out how we get cases resolved efficiently given many factors, one of which is ESI which has to be dealt with in the case.

I don’t pretend to believe we’ve solved the problem. I think you’re right – we don’t really foresee the extent of the problem and it’s something that the courts are going to have to adapt to. But our intent is to at least make some progress on dealing with ESI through these amendments.

“I don’t pretend to believe we’ve solved the problem. We don’t really foresee the extent of the problem.”

Logikcull: Judge, what’s your sense of how well technology — and I’m now talking about the legal provider side, or e-discovery vendor side — has developed and been adopted by practitioners to counter these rising data volumes? I’m particularly interested in knowing your thoughts on, if you’ve even seen it, the utility of predictive coding… The field has put a ton of faith into that process and others, and into technology in general — and it seems to me, anyway, that the results have been somewhat underwhelming. Do you have an opinion on that?

Campbell: I’ve got some thoughts. I don’t know if they’re mature enough to call them an opinion [laughing]… I think it’s a reality that, going forward, we’re going to have to find technological solutions to this growth of ESI. The reality is that the old model of having lawyers or paralegals review every document that’s produced is not going to work. It just can’t work in a world where you’re dealing with millions of documents. We either incur enormous expenses continuing that model; or we surrender and produce everything — which I don’t think lawyers are going to do; or we need to find a technological way to winnow down the ESI to a manageable size.

So my view is, whether it can accurately be said that, today, technology is solving the problem, ultimately it’s going to have to solve the problem, because I don’t think the court system and lawyers are going to be able to continue dealing with it the way we dealt with paper evidence.

I will also say that I’m finding in my cases that predictive coding or technology-assisted review is rarely proposed by the parties and rarely used. I think that will change over time. I’ve seen more of it over the last few years. But it is still used in a very small percentage of the cases. And I am surprised at the number of big document cases where the parties do not use it, even when I suggest it. They instead prefer keyword searches.

I think part of that is the need to educate the bar through sophisticated litigators as to what the technology can do, and it is my hope that the predictions will prove true that it really can do this more quickly and more accurately than people.

But I’m not seeing it used widely in my cases, and, you know, Phoenix is the fourth or fifth biggest city in the United States. So I think we have our share of complex cases, and yet it is still a rare commodity in my cases.

“Whether it can accurately be said that, today, technology is solving the problem, ultimately it’s going to have to solve the problem, because I don’t think the court system and lawyers are going to be able to continue dealing with it the way we dealt with paper evidence.”

Logikcull: You mentioned the emphasis in the rule changes on making judges more proactive case managers. Judge [Paul] Grimm [of the federal District Court in Maryland], among others, has been pretty vocal about the idea that judges do not traditionally view themselves as case managers, but as dispute resolutionists. What’s your assessment of how well the judiciary will be able to take a more proactive role in facilitating some of these e-discovery issues?

Campbell: I have no illusion that simply re-writing the rules is going to transform judges from passive to active case managers. It clearly won’t. And as you know, the idea of active case management has been in the Federal Rules since 1983, when Rule 16 was revised to create an active role for judges. So we have been of the view that these rule changes need to be accompanied by a very significant education effort to encourage judges, in particular, to be more active case managers. Many are. But Judge Grimm is right that many still are not.

There are a number of steps we’ve taken. Members of the committee have written articles. We’ve created videos that are now on the Federal Judicial Center website that have been sent to every federal judge in the country explaining the rule changes. The committee has written letters to every chief district judge and every chief circuit judge in the country asking that they include the new rule amendments in district and circuit conferences next year to educate judges and lawyers about them. We’ve compiled materials that are available for judges on the FJC website — articles, PowerPoints and other things. And the Federal Judicial Center, which is the entity that trains judges, is intending to do more active training of judges. We’re hoping that that push, along with the rule changes, will bring about a behavior change on the part of us judges. Whether or not it works, we’ll have to see over time. But I’m hoping, at least with many judges, it will produce more active case management.

Logikcull: To stay on this theme of education, but now turning to educating the bar and practitioners… I get the sense, just as an outside observer, that there is an education gap. You have some small percentage of highly knowledgeable people who are technically competent who typically, though certainly not always, come from the largest law firms and largest corporations. And then you have some large remnant that tends to not understand this stuff at all. You’re at the federal court level. Is that what you’re seeing? Or are you seeing a general rising of education?

Campbell: Well I certainly see lawyers who understand the rules and are dealing with ESI better than others do. But I don’t think I would say that you find all of the best prepared lawyers in the large law firms or large corporations. I see many lawyers who are sole practitioners or who are in small firms or government attorneys who are right on top of ESI and the rules.

But there’s no doubt that many are not. I think that is changing. I alluded to it a moment ago, I’m seeing more and more — although it’s still a fairly small percentage — thinking about things like technology-assisted review. And I’m definitely seeing more who are dealing with ESI up front and talking about it at the Rule 26(f) conference. I’m hoping that the publicity that occurs in connection with these rules amendments — and there is a fair amount of publicity going on through various bar groups — will educate lawyers about these ESI issues in a way that they haven’t been before. As I said earlier, we have to deal with ESI in the federal courts. I think everybody involved is becoming more and more aware of it — and we’ll see lawyers become more sophisticated over time.

“I’m seeing more and more (lawyers) — although it’s still a fairly small percentage — thinking about things like technology-assisted review.”

Logikcull: Some practitioners see the expense associated with e-discovery as an access to justice problem. Judge John Facciola has gone so far as to say that e-discovery is contributing to making the federal court system a “playground for the rich.” Is that a sentiment that you share?

Campbell: I don’t think I would call it a “playground for the rich,” but I absolutely agree that too many people cannot afford to litigate in federal court. I do think the cost of federal litigation makes it unavailable to the average citizen. And I see many of them who are representing themselves struggling to handle a case because they can’t get a lawyer to take it because it doesn’t have enough money at stake. I think that’s a problem.

“I do think the cost of federal litigation makes it unavailable to the average citizen. And I see many of them who are representing themselves struggling to handle a case because they can’t get a lawyer to take it because it doesn’t have enough money at stake.”

That’s one of the problems we talked about at the 2010 conference that I mentioned. Part of our intent in putting the proportionality idea into the new rules and trying to get judges to actively manage cases more efficiently from the beginning, and we hope cutting down the side litigation over the loss of ESI, is to reign in the cost of discovery. There are other things that the Civil Rules Committee and other Judicial Conference committees are looking at to try to make civil litigation less expensive.

But in my view, it absolutely is a problem, and one that we need to work hard as a federal judiciary to solve.