White House announces commitments by leading companies to responsible development of AI technology

Last month, the risks of relying on artificial intelligence (AI) systems hit home for the legal profession when a brief filed in a New York federal court was found to rely on a series of fictitious, AI-generated case citations. This incident coincided with a broader surge in public interest in generative AI systems that can produce new content - whether in the form of text, images, video, music, or otherwise - in response to user prompts. While these models have shown amazing creative and commercial potential across a range of applications, they have also gained the attention of government officials who see potential risks and unanswered questions about how this technology may be used.

In response to the public's concerns, the White House today announced that several major technology companies have committed to implementing safeguards to ensure the safe development of generative AI systems. Among the safeguards are commitments to develop watermarks on AI-generated content to reduce the risk of fraud and deception, to internally and externally test generative AI systems' safety before public release, and to invest in research on how AI models can protect privacy and safeguard against potential bias and discrimination. The announcement was not solely focused on risk mitigation; the commitments showed the broad potential for AI to transform sectors, and the companies committed to developing and deploying AI systems to help address some of society's greatest challenges, including cancer prevention and mitigating climate change. In his remarks, President Biden called these commitments "real" and "concrete."

These commitments are the latest in a series of moves made by an administration that has shown an eagerness to take action on AI policy. In October 2022, the White House released a Blueprint for an AI Bill of Rights setting out principles to guide the design and deployment of AI systems that respect Americans' rights, and in May of this year, the Office of Science and Technology Policy released a Request for Information seeking input on priorities that should be incorporated into a National AI Strategy. And in one of the most concrete actions taken to date, the National Institute for Standards and Technology has developed a Risk Management Framework that organizations can use to develop "trustworthy AI models."

But none of these actions, including today's commitments, amount to a comprehensive approach to federal regulation of generative AI that would broadly impact the private sector. While there have been a number of rulemaking efforts in related areas - such as by the Federal Trade Commission (FTC) on commercial surveillance and data privacy and the Securities and Exchange Commission on cybersecurity - the federal government has not outlined a whole-of-government approach to AI regulation. Recently, the Federal Elections Commission even rejected a call to develop a rule prohibiting the use of deceptive generative AI content in campaign ads. And while some agencies and regulators, such as the FTC, are tracking developments, neither the executive branch nor Congress has clearly articulated a vision for comprehensive regulation for this rapidly developing space.

This cautious approach to regulation stands in stark contrast to the rapid push toward regulation in other parts of the world, including in particular the European Union, which is in the final stages of a multi-step process to adopt the AI Act. If enacted, the AI Act would constitute a European Union-wide regulation that comprehensively regulates and restrains AI, and expressly contemplates generative AI systems. How this plays out in practice, including the impacts on innovation, product availability, and intersection with other European Union legal regimes, including the General Data Protection Regulation (GDPR), remains to be seen. Meanwhile, across the Pacific, China has already issued interim regulations governing generative AI systems that are available to the Chinese public.

Nevertheless, today's announcements are meaningful and will no doubt inform lawmakers on Capitol Hill as they consider how to balance innovation and progress with the principles of safety, security, and trust that leaders of some of the most prominent AI companies agree are critical to the successful growth of the sector and the achievement of generative AI's potential benefits to society.

Indeed, legislation may be on the horizon. Senate Majority Leader Chuck Schumer is leading a bipartisan group of senators to craft AI legislation that is anchored around security, accountability, and explainability that could be introduced early next year, following the conclusion of a series of forums on AI this fall. In recent months, Congress has also held a variety of hearings on the potential risks of AI, both in the Republican controlled house and the Democratic controlled Senate. In remarks today, President Biden expressed support for Senator Schumer and other members of Congress who were working to build bipartisan legislation to regulate generative AI.

For now, today's commitments set baseline expectations for how companies develop - and, to some extent, deploy - AI tools. As the global race for leadership in generative AI heats up, let's hope that the United States puts its best and brightest minds to work on the vital task of ensuring thoughtful and balanced AI oversight.

The opinions expressed in this article are those of the authors and do not necessarily reflect the views of the firm or its clients.