Artificial Intelligence (AI) has become a crucial tool for organizations in various sectors, particularly in the generation of content and code by generative AI systems such as ChatGPT, GitHub Copilot, AlphaCode, Bard and DALL-E, among other tools. As the promise of incorporating these generative tools in the corporate setting is all but assured in the near term, there are a number of risks that need to be minimized as companies more forward. In particular, as AI applications grow increasingly sophisticated, they raise concerns with several forms of intellectual property (IP), such as patents, copyrights, and trade secrets. This article aims to discuss these issues and provide a sample company policy for using AI-generated content such as software code.
The story seems to unfold the same way every time, whether the actor is a high-level departing employee or a customer or business partner. When sharing confidential information in a long-term relationship results in the release of a similar product by the recipient, the reaction is a claim of theft, laced with accusations of treachery and betrayal. And the response is equally strong: “no, I did this on my own”; in legal terms, “I engaged in ‘independent development.’” Strictly speaking, this means that the development of the new product was accomplished “independently” of the information shared in the confidential relationship. As a practical matter, this can be difficult to prove. Once you have been exposed to the secret process or design, or other related information, how do you demonstrate that your work was entirely your own?
As I described in my previous article about the protocol for clean room product development, there are often situations where a company has come into contact with intellectual property that it cannot allow to spread to a product in development. In that article, I described the concepts of a clean room. In this article, I describe the practical implementation details. Once a clean room protocol has been set up, the following sections describe the recommended procedure to implement.
In January of this year, the Federal Trade Commission (FTC) proposed a new rule that would ban employers from using noncompete clauses for their employees. In an announcement, the FTC said that the use of noncompete clauses is “a widespread and often exploitative practice that suppresses wages, hampers innovation, and blocks entrepreneurs from starting new businesses.” The agency estimated the new rule could increase wages by $300 billion a year, as firms would be encouraged to do more to keep their workers. The proposed rule change was opened for public comment in January, and the deadline for submissions was extended from March 20 to April 19 in early March. As of April 18, the Regulations.gov website indicated that 24,259 comments had been received and 14,946 posted. With the comment period coming to a close this week, the U.S. Chamber of Commerce has weighed in, urging April Tabor, FTC Secretary to withdraw the proposed rule.
It was a hot August afternoon in 1984, and I had just finished testifying to the California Senate committee considering a new law, the Uniform Trade Secrets Act (UTSA). I had been sent to Sacramento to support this legislation, which was supposed to provide a “uniform” standard among the states. But some lawyers from the State Bar were pushing for changes that I thought might cause problems. One of these was to remove the requirement that a trade secret owner prove the information was not “readily ascertainable.” If you’re still reading, well done! You’ve demonstrated your intellectual curiosity. Please keep going; I promise this will not be a dry, academic rant about something that can’t possibly matter to you. Instead, this is a story about the unintended consequences of casual law-making and the ways that courts can amplify those effects without really understanding what they’re doing.
It’s getting pretty rough out there for employers who want to control their employees’ behavior. Think back to March 2020, when the pandemic was just beginning and we took a look at this new phenomenon of widespread remote work. We imagined managers wistfully recalling the Renaissance, when artisans could be imprisoned, or even threatened with death, to make sure they didn’t breach confidence. Well, in modern times at least, companies can use noncompete agreements with departing employees to avoid messy and unpredictable litigation over trade secrets. Maybe not for long. As we learned last month, the FTC is on the warpath about noncompetes, and it may not be long before the entire country is forced to emulate California and just do without. Whatever happens with the FTC proposal, it’s pretty clear that noncompetes are also under attack by the states, where new laws limit their effectiveness.
In January 2023, the Federal Trade Commission (FTC) unveiled a proposed ban on non-compete clauses that prohibits employees from joining or forming competitive firms following the termination of their employment. According to the FTC, non-compete clauses unfairly and unnecessarily stifle employees’ ability to pursue better employment opportunities. While this criticism may ring true in the case of lower-wage workers, such as restaurant and warehouse employees, even the staunchest critics of non-compete clauses will typically acknowledge that they can — and often do — play a legitimate role in the protection of trade secrets. This is why the FTC’s proposed rulemaking is causing consternation in the intellectual property community.
When Adam Smith spoke about an “invisible hand,” he was talking about a good thing – the way that free markets harness the laws of competition, supply and demand and self-interest to improve the economy. But he also could have been thinking of another law. The law of unintended consequences: that actions of people, and especially of governments, always have unanticipated effects. Sometimes these effects can be perverse, reflecting a profound failure of “second-order thinking” (in other words, thinking ahead about “how could this possibly go wrong?”). On January 5, 2023 – a day that may go down in IP infamy – we saw two bold actions. First, the “Protecting American IP Act” became law; and second, the Federal Trade Commission (FTC) proposed a new rule that would invalidate noncompete agreements across the United States. But wait, you might say, that actually sounds great! What’s the problem with protecting American IP, and making the rest of the country join California in unleashing talent to go where it likes? Well, don’t be too hasty. Stay with me on this, and you will see just how shortsighted our government can be.
Trade secret jurisprudence, originally conceived in the common law of torts as a way to enforce confidential relationships, now has a sharper focus directed at the property interest of businesses in the data that forms the major portion of their asset base. In the process, trade secrets have taken their place of respect alongside the “registered rights” of patents, copyrights, trademarks and designs. But just because we now enjoy statutory guidance through the Uniform Trade Secrets Act (“UTSA”), enacted with some variations in every state but New York, and national uniformity in federal courts through the Defend Trade Secrets Act of 2016 (“DTSA”), the law continues to evolve much as it did a century ago—that is, through the opinions of judges deciding individual cases on their facts.
The U.S. District Court for the Southern District of New York last week dismissed a trade secrets lawsuit against International Business Machines Corp (IBM) and IBM China by Beijing Neu Cloud Oriental System Technology Co. The Chinese firm alleged that IBM stole trade secrets from its joint venture in order to sell IBM products to the Chinese market. IBM China and Beijing Teamsun Technology Co. originally formed Beijing Neu Cloud in 2014 as a joint venture to distribute IBM technology in China. But in a 2021 complaint, Beijing Neu Cloud alleged that IBM induced “Neu Cloud and its majority owner through later-breached contracts to expend resources and provide IBM with access to sensitive, confidential customer information, which IBM then secretly used to create competing ventures in China.”
It was February 2017 when Waymo, Google’s self-driving car unit, sued Uber in what would become the biggest trade secret case of the century. Waymo alleged that its former manager, Anthony Levandowski, had organized a competing company while still at Waymo, and before leaving had downloaded 14,000 confidential documents. As it turned out, Uber had known about this when it agreed to pay $680 million for Levandowski’s brand new startup; and we’ve already looked at how the hubris of that hasty transaction provides lessons for hiring in new markets driven by emerging technology.
National Science Foundation research shows that many R&D-oriented companies believe that trade secrets are more important than patents and copyrights. How did this happen, and why are trade secrets growing in importance? Bruce Berman, host of the “Understanding IP Matters” podcast, sought out trade secrets expert Jim Pooley to find out why. Pooley is the world’s foremost expert on trade secrets, a mysterious area of the law that has been the focus of employer disputes. A successful Silicon Valley trial lawyer, Pooley served for five years as Deputy Director General of the World Intellectual Property Organization (WIPO) in Geneva, Switzerland. His commentary pieces on the controversial COVID vaccine patent waiver and other topics have appeared in The Wall Street Journal and The Financial Times, and he is a regular contributor on IPWatchdog.
On September 2, the U.S. Court of Appeals for the First Circuit issued a decision in Amyndas Pharmaceuticals, S.A. v. Zealand Pharma AS affirming the District of Massachusetts’ decision to dismiss trade secret misappropriation claims between former drug development partners. However, the First Circuit found that the district court abused its discretion in denying Amyndas’ motion to file an amended complaint and vacated the dismissal of trade secret claims against Zealand’s U.S. subsidiary.
The U.S. Court of Appeals for the Third Circuit on Monday said in a precedential decision that Jiangsu Tie Mao Glass Co. Ltd. (TMG) should have shown up sooner in a trade secrets misappropriation lawsuit brought against it by PPG Industries if it wanted to have a chance at winning. But by failing to enter the litigation until after PPG asked the district court to enter default judgment and award damages for unjust enrichment, “its protestations were and are too little and much too late,” said the appellate court.
An “NNN” agreement is short for Non-Disclosure/Non-Use/Non-Circumvention agreement, which means the information cannot be shared with anyone, it cannot be used in any way, and “behind-the-back” or design around tactics are forbidden. In recent years, signing NNN agreements has become widely adopted and is now the standard initial step in dealings with Chinese companies, particularly original equipment manufacturers (OEMs). An NNN Agreement is much more than just a Non-Disclosure Agreement (NDA). An NDA focuses narrowly on preventing secret information from being revealed to a third party or to the public, which is not sufficient for OEMs in China. In contrast, an NNN agreement not only contains confidentiality provisions, but also prevents misuse of confidential information.