Skip to content

Checkpoints Ahead: Parsing the AI Risk Environment

Since generative AI burst into the mainstream, companies have raced to capitalize on its extraordinary promise. But as with any technological frontier, this promise does not come without risks, and companies can expect to encounter them with greater frequency as AI’s role in the economy continues to grow.

Vinson & Elkins works closely with clients to identify, manage, and mitigate AI’s many emerging risks. Below, four of our partners share their view of the AI risk environment in their practice area.

Antitrust

“There is perhaps no other industry generating the same level of interest at competition agencies today. US and European enforcers are already looking at whether several announced partnerships involving GenAI services raise antitrust concerns, and we expect enforcers to carefully scrutinize distribution strategies for these services. As GenAI developers seek out new sources of training data, investigations premised on concerns around the acquisition of this data, including from companies that are not in the tech sector, will become more frequent.

We are also seeing allegations of algorithmic collusion as an emerging issue. This involves use of a common AI tool by competitors in the same industry to allegedly collude on pricing or some other form of competition. With antitrust regulators worldwide initiating market studies on GenAI, we anticipate more investigations and acquisitions in this sector in the near future.”

Darren Tucker

Darren Tucker
Partner, Antitrust

Energy Regulation

“As AI technology advances, the demand for power surges, yet the supply remains constrained. Conventional energy sources like fossil fuels are dwindling, prompting a critical need for massive solar, wind, and nuclear power installations by 2030. However, infrastructure challenges compound the issue. The existing transmission grid is not designed to handle this new load and generation. And it can take up to five years, if not longer, to construct new lines.

So, solving the supply-demand imbalance created by the AI boom is not as simple as building new generating facilities, which is no easy task itself and involves many regulatory hurdles. Rather, AI will require a massive build-out of the electric grid, which is still heavily regulated. Moreover, the lack of energy-storage solutions threatens stability as renewable power is added. In this complex landscape, AI acts as a disrupter, intensifying the demand for innovative energy solutions and overcoming regulatory hurdles.”

Jeffrey Jakubiak

Jeffrey Jakubiak
Partner, Energy Regulation

Government Investigations

“In March, Deputy Attorney General Lisa Monaco made an important statement highlighting the Department of Justice’s efforts to increase enforcement and seek harsher penalties for criminal offenses that utilize AI, particularly in market manipulation, price fixing, and fraud cases. Our clients need to be mindful of the need to ensure that there is sufficient oversight of how AI is employed within their companies, so that the positive uses of AI can be harnessed and controls are in place to guard against potential misuse.

It is now more important than ever for companies to be able to defend their use of AI and demonstrate that they have adequate resources committed to ensuring that their use of AI is compliant and plays a lawful and constructive role in company operations.”

Ephraim (Fry) Wernick 4x5

Ephraim (Fry) Wernick
Partner, White Collar Crime and Government Investigations

Intellectual Property

“The current utilization of AI focuses primarily on enhancing efficiency and automation. However, a significant shift occurs when AI steps into the realm of intellectual property, particularly as a potential inventor. Earlier this year, the US Patent and Trademark Office issued guidance emphasizing the necessity of a human guiding AI in the inventive process. This aligns with a recent court decision confirming that an inventor must be a person, not AI. This decision underscores the challenges in distinguishing between the role of human oversight and the autonomous actions of AI in the creation of new inventions.

If there aren’t changes to the Patent Act to allow AI systems to be inventors, there could be a shift to innovators using trade-secret law for protection instead of patents. To get a patent, you must fully disclose your invention, so that others can build upon it after the term expires. Why would someone fully disclose their invention to the Patent and Trademark Office and risk not getting a patent? This would be a huge, fundamental shift in technology management and strategy.

The problem with using trade-secret law for protection is that it must be kept a secret. If someone can reverse engineer your system, it’s not a secret and they can use it. People will use AI systems to do that reverse engineering, so the incentives and economics to innovate will change. The real race will be to have the most powerful AI systems that can solve problems in such a way that other AI systems either don’t understand or cannot reverse engineer or deconstruct them. It is critical that companies fully consider and understand this dynamic, complex situation to determine how best to protect their most important technology.

Eric Klein

Eric Klein
Partner, Intellectual Property

This information is provided by Vinson & Elkins LLP for educational and informational purposes only and is not intended, nor should it be construed, as legal advice.