Monday, April 27, 2015

I Cannot Stop Writing about Sergey Aleynikov

When Sergey Aleynikov sued the FBI agents responsible for his 2009 arrest, his journey through the court system officially became what we call "Batshit F**king Crazy." After all, this guy appears not to have injured (physically, emotionally, or financially) anyone. His ex-employer, Goldman Sachs Group, seems to have survived his resignation just fine.

Aleynikov, for those who still don't know, is a former Goldman Sachs computer programmer who left to join Teza Technologies in Chicago. However, he made some mistakes when leaving. He apparently transferred computer data from Goldman's high-frequency trading platform to a remote server in Germany, allegedly for use at Teza. (The conclusion may be dubious, for it's unclear what value a fraction of HFT code would have to a third-party, but that's for Aleynikov's highly skilled lawyer to argue - not me.)

To be sure, many a trade-secret defendant makes a similar mistake. But not everyone works for a Goldman Sachs. Aleynikov has turned the tables on Goldman before, and with his most recent lawsuit, he now seeks to turn the tables on those who Goldman allegedly conscripted - the officers who arrested him after he took the computer data.

In most cases, a fact set like this gives way to a rather bland cease-and-desist letter (or more accurately, a cease-desist-and-return letter). Perhaps the parties fight and end up at an injunction hearing. Even then, 90+ percent of such cases settle fairly quickly - often without payment of any money and relatively minor conduct restrictions agreed-upon.

If only it were so easy for Sergey.

To put this in context, we're talking about a guy who:

  1. Provoked an Illinois lawsuit involving Aleynikov's would-be employer, Teza Technologies, LLC, which resulted in a significant and somewhat controversial non-compete rule in our state;
  2. Was convicted in federal court under the National Stolen Property Act and the Economic Espionage Act for transferring some of Goldman's high-frequency trading source code to a remote server in Germany;
  3. Served 51 months in federal prison;
  4. Had his conviction reversed by the Second Circuit Court of Appeals;
  5. Was released from federal prison the same day as his lawyer's oral argument in the Second Circuit;
  6. Effectively encouraged (through the reversal of his conviction) Congress to amend the Economic Espionage Act for the purpose of redefining what type of conduct violate the statute, on a vote of (hold your breath) 388-4;
  7. Was re-arrested and charged under New York state law for theft of intellectual property;
  8. Sued Goldman Sachs to have it advance his legal fees in the New York state proceeding;
  9. Won that advancement action on summary judgment;
  10. Had his advancement case reversed, over a dissent, by the Third Circuit Court of Appeals;
  11. Served as the inspiration for Michael Lewis' phenomenal book Flash Boys, which itself delayed the IPO of Virtu Financial;
  12. Went to trial on the state intellectual property theft charges last week, with a jury still (as of this writing) deliberating over his fate (the trial judge appeared skeptical though); and
  13. (as noted below) Sued FBI agents in a so-called Bivens suit for violating his constitutional rights in arresting him without probable cause back in 2009.

In my opinion, Aleynikov's Bivens case may run directly into the defense of qualified immunity since he must show that the officers violated clearly established law in arresting him back in 2009. Taking the allegations as true, it's hard to see how the law back in 2009 was clearly established such that the federal officers responsible for his arrest now must account for legal damages. After all, Congress did amend the Economic Espionage Act after the Second Circuit reversed Aleynikov's conviction. And the case law on EEA prosecutions is (and was) sparse at best.

The most interesting aspect of the Bivens case is that Aleynikov's essentially treats the FBI as Goldman's agents, willing to do Goldman's bidding in a criminal case. In truth, he needs to make those allegations to get around qualified immunity. His case isn't frivolous, but he seems to be swimming uphill.

Strictly from a lawyer's perspective, even if the Bivens suit fails, I again have to commend Aleynikov's lawyer, Kevin Marino. From his work in getting the conviction overturned to the advancement proceeding in New Jersey, he has achieved simply extraordinary results for Aleynikov and has managed, at times, to outwit the Manhattan DA, the US Attorney, and Goldman itself.

The complaint he filed in the Bivens case is at times bombastic and inflammatory, but it's pretty damned effective and a great read. He knows precisely what he's doing. Whatever happens, and against all odds, Aleynikov fought back and then some. And that too is pretty rare to see.

Friday, April 17, 2015

The Golden Rule: Ninth Circuit Issues Odd, Unclear Opinion on No-Employment Clauses

My colleague Robert Milligan at Seyfarth Shaw predictably beat me to the punch in commenting on the Ninth Circuit's recent opinion in Golden v. Cal. Emergency Phys. Med. Group, which is embedded below. But at the risk of repeating his fine and thorough discussion of the case, it's potentially an important enough development for me to comment on as well.

The court held that a "no-employment" clause in a settlement agreement could constitute an invalid restraint of trade under Section 16600 of California's Business & Professions Code. No-employment clauses are neither unusual nor controversial, and they often appear in settlement agreements arising out of employment lawsuits.

Dr. Golden appeared to have settled an employment dispute with California Emergency Physicians Medical Group, which is apparently a large consortium of physicians that staffs ERs in California. As part of the settlement, Dr. Golden had to agree to a no-employment clause with CEP and any facility that CEP may own or with which it may contract in the future. It also gave CEP the right to terminate Dr. Golden from a facility if CEP acquired it in the future.

Dr. Golden, though, balked at signing the settlement agreement on account of the no-employment provision, arguing that it violated Section 16600 of the Code. Courts have interpreted that provision broadly, invalidating virtually all forms of employee covenants not to compete. However, the no-employment clause is not really a covenant not to compete. It arises in the context of a settlement, rather than as part of any hiring or other employment condition.

The Ninth Circuit sided with Dr. Golden and remanded the case back to the district court for fact-finding. Importantly, the court did not find that the no-employment clause was an impermissible restraint of trade under California law. Rather, it found the district court erred by categorically holding that it wasn't. Put simply, a "no-employment" clause can implicate Section 16600 and that section is not limited to a traditional non-compete. Therefore, a remand was necessary so the district could "conduct further fact-finding as it deems prudent."

That's where the case is a bit of an oddity (shocking for the Ninth Circuit). As the dissent points out, fact-finding over what, exactly? The agreement is what it is, and there was nothing immediately that impacted Dr. Golden's employment. So what are the pertinent facts?

Perhaps the court is saying that the trial judge must determine how significant a player CEP is in the regional ER staffing business and how its market power could affect Dr. Golden. If so, this starts to sound like a mini-antitrust trial in the context of an employment settlement. Lovely. Or maybe the trial judge will examine CEP's expansion strategy to see how big it could get. Without question, CEP will love to reveal those cards. Or should a labor economist testify about the potential future impact on Dr. Golden's ability to practice medicine and what opportunities he might have to forego? That seems more valid, but who pays for that? Does Dr. Golden's settlement cost him expert witness fees, all of a sudden? Ultimately, the case isn't clear, and there's little guidance for the district court on remand.

The impact of this case beyond the facts is unclear. In my opinion, lawyers may make more of this than they actually should. A typical no-hire clause in an employment settlement agreement will limit an employee's ability to work for that employer. It's unclear that many employees would want to contest that, and Dr. Golden is in somewhat of a unique position given the size and scope of the consortium he sued. Too, Dr. Golden and his lawyer appear to have had some sort of a falling out, and Dr. Golden held up the settlement on account of this provision. This doesn't happen all that often, so it's pretty likely that even if these clauses are unenforceable no employee is going to torpedo a settlement because of its inclusion.

Almost certainly, employers will want to add some representation or warranty in a no-hire clause from the employee that he does not believe the clause will constitute a restraint of substantial character on his trade or profession. That may help, but it's not failsafe. Or, employers could leave the clause out entirely and just not hire the person back. I don't see what's so wrong with that solution. Seems simple and less of a lawyer quandary.

Friday, April 10, 2015

New Ruling on Computer Fraud and Abuse Act Illustrates Division in Eleventh Circuit

The reach of the federal Computer Fraud and Abuse Act continues to divide courts and commentators. The friction spills over into statutory language that seemingly is pretty clear: one cannot access a protected computer to obtain information in a manner that exceeds the user's authorized access.

The phrase "exceeds authorized access" has yielded a split among many federal courts. On one side sits the pro-employer theory: that an employee cannot access information for reasons inconsistent with the employer's interest. On the other side, employees argue that the CFAA's definition does not turn on subjective intent as to use and that the statute only bars the accessing of information that the employer did not allow.

The Fifth and Seventh Circuits have pretty well-established rules that fall on the pro-employer side, though the rationales are not identical. The Fourth and Ninth are decidedly pro-employee and construe the CFAA much more narrowly.

The Eleventh Circuit's history on this is not straightforward.

Courts within the Eleventh Circuit (that is, district courts in Florida, Georgia, and Alabama) are divided on how to interpret and apply the CFAA. Some follow the pro-employee construction. Others disagree and side with the Fourth and Ninth Circuit line of authority. The problem seems to be a criminal case from the Eleventh Circuit itself, United States v. Rodriguez, which arguably requires a pro-employer construction of the CFAA.

Illustrating this divide is Enhanced Recovery Co., LLC v. Frady (opinion embedded below), which canvasses the law and decides not to follow Rodriguez - arguably binding precedent in the Eleventh Circuit. Frady involves perhaps the most common, ubiquitous CFAA fact pattern in employee mobility cases. The allegations hinged on an employee's act in preparing to compete by sending corporate documents from a company e-mail account to a personal web-based e-mail. Under the employer's theory, this transmission of e-mails is antagonistic to its business interests and constitutes the use of company information in an unauthorized manner.

As with many close CFAA cases, the fact pattern in Frady adds a twist because the employer had fairly clear policies against non-disclosure of information and a requirement to return company materials at the termination of employment. One reasonably could make the argument that these policies at least put the employee on notice that accessing corporate information for a prohibited use exceeded any authorization to use that information.

The district court in Frady was not persuaded and held this wasn't enough. In doing so, it seemed to tread a very narrow path through the Rodriguez case, which seems on first read to warrant a broader application of the CFAA.

My personal opinion is that something else must be going on here. I have seen judges walk this tightrope before, and I believe that regardless of any formal legal analysis, many remain troubled by the extension of the CFAA well beyond its intended purpose to punish hackers criminally. Courts, of course, acknowledge this but judges must default back to the statutory text to see whether conduct falls within the statutory proscription. And they are certainly bound by circuit law. In the CFAA arena, the distinctions that judges are drawing appear to be so fine and nuanced as to be meaningless.

This is not to say that the court in Frady got the result wrong as a normative matter, because the CFAA is overused. It is, to be sure, not a trade secrets law substitute - though many attorneys perceive that it is. And if the Obama Administration has its way, this divide over the meaning of authorized access will get cleaned up in a way that (surprisingly) favors the employer. Ultimately, I think this too is wrong. The existing legal framework under state law works just fine, and a federal trade secrets act may be okay too (though I am more lukewarm on that than others in my profession).

However, extending the CFAA and criminalizing conduct that doesn't even violate state trade secret law is not the right way to go about protecting intellectual property. In the employment realm, it remains a terrible fit. How judges get there is concerning, but if they agree, then I'm okay with a formal legal analysis that leaves me scratching my head.

Thursday, April 9, 2015

Streisand Effect Bedamned: Jimmy John's "Prevails" In Class Action Non-Compete Dispute

Quite possibly because my 3-year old loves Jimmy John's sandwiches, I have a profound affinity for the place. It's a minimum once-per-week destination. Plus, I went to law school in Champaign, Illinois, where Jimmy John's is somewhat of a revered institution.

Even I have to admit that I was a bit discouraged by reading the negative publicity surrounding Jimmy John's standard "Confidentiality and Non-Competition Agreement." It appears to apply to your average sandwich-maker and prevent him from jumping ship to Subway (but not Chipotle!). The mind reels at the thought of a stoned 19-year old testifying on the stand about the proprietary nature of applying roast beef between lettuce and mayo.

But the lawsuit - which sought a declaratory judgment that the non-compete was invalid - seemed doomed from the start. And yesterday, Judge Charles Kocoras dismissed the putative class action in Brunner v. Jimmy John's LLC, 2015 U.S. Dist. LEXIS 46018 (N.D. Ill. Apr. 8, 2015). The rationale was lack of standing, a somewhat formal legalist concept that the plaintiffs did not present enough of a concrete case to have a court decide.

In other words, the mere existence of the agreement was not enough to create a controversy that a court is capable of deciding. The putative class plaintiffs alleged - and I believe them - that they didn't really know what the non-compete meant or how Jimmy John's might apply or enforce it. This, by itself, did not create enough of an apprehension that a lawsuit was imminent. Put another way, the dispute was too hypothetical.

Jimmy John's also offered, sensibly in my view, affidavit testimony that revealed an intent not to enforce the non-compete against the putative class plaintiffs in the future. This, alone, may not be enough to demonstrate the lack of a case or controversy, but it sure helped Jimmy John's here.

(As an aside, the standing rules in federal court are a bit more stringent than in many state courts. Having had first-hand experience dealing with this, it's pretty clear that any type of a non-compete declaratory judgment case in federal court runs the possibility of being dismissed for lack of standing.)

Despite the win, this all feels like somewhat of a loss. The publicity on The Huffington Post and elsewhere surrounding the corporate mandate certainly damaged Jimmy John's and created a storm of negative publicity. just removed its non-compete for hourly workers, a tactic which resembles Jimmy John's representation that it won't enforce the agreement.

This whole affair reminds me of the Streisand Effect, although the analogy is imperfect. The thinking is that the attempt to remove information can have the effect of promoting it. This unintended consequence often appears in defamation cases, where a plaintiff attacks a disparaging comment only to put his reputation on trial.

The lawsuits and publicity surrounding the Jimmy John's and non-compete work a bit in reverse, but the concept is the same. Even a win may be a defeat.

Friday, April 3, 2015

Michigan Legislator Introduces Bill to Ban Non-Compete Agreements

As reported recently by Crain's Detroit Business, State Rep. Peter Lucido introduced House Bill 4198, which would ban employee non-compete agreements in the State of Michigan. The bill, which has no co-sponsors (and which has little chance of passage), proposes to amend MCL 445.774a, the state code section pertaining to covenants not to compete.

As amended, the proposal would allow for enforcement of covenants incidental to the purchase of goodwill. However, it specifically would void any restrictive covenant entered into between employer and employee.

Michigan has a somewhat circuitous history with regard to non-competes. Thirty years ago, employment non-competes largely were void, subject to some narrow exceptions. In 1985, however, the legislature reformed non-compete law to bring the state largely within the majority of other states, effectively allowing them as long as the terms are reasonable and are designed to protect a valid competitive business interest. Lucido would take the law back to its pre-1985 status.

Matt Marx, Deborah Strumsky, and Lee Fleming published a well-known Management Sciences article in 2009, which tested the effect of the law's evolution on the mobility of Michigan employees. The study concluded that the change in Michigan law positively correlated to a brain drain, as star inventors left Michigan for states with more lax non-compete enforcement policy. In many respects, the Marx-Fleming study is the intellectual forbear to Orly Lobel's book Talent Wants to Be Free (Yale Univ. Press 2013).

In the Crain's article, attorney Bernard Fuhs takes issue with Rep. Lucido's rationale for getting rid of non-competes. Confidentiality Agreements, Lucido says, are sufficient to protect intellectual property. Fuhs offers the best retort: it's virtually impossible to police compliance with a non-disclosure agreement absent evidence from a whistleblower or through sheer dumb luck.

Non-competes provide a more objective means to ensure protection of intellectual property, even if they are somewhat overinclusive. By "overinclusive," I mean that non-competes impede a flow of knowledge or value that can be restricted only incidentally and that may not provide any competitive advantage at all.

The entire rationale for enforcing non-competes is that the economic benefit from preventing disclosure of proprietary information exceeds the social cost in preventing use of non-proprietary spillover knowledge .