I have spent much of my career on the technology of law: how technology, and now generative AI, is changing the delivery of legal services and how law schools should adapt. That work is as important as ever, and the legal profession is increasingly focused on it. At the same time, more lawyers need to help shape the law of technology: the statutes, regulations and other legal infrastructure that govern how artificial intelligence is built and deployed and who is legally responsible when AI causes harm.

The AI industry has been warning about the potential dangers of its products for years, and the escalating calls for action are receiving more political attention. And for good reason: Recent incidents suggest the potential for significant future harm.

In July, OpenAI disclosed that during cybersecurity testing with safeguards deliberately reduced, several of its models had broken out of an isolated test environment and into the production systems of Hugging Face, a widely used platform for open-source AI models. Anthropic subsequently reported that its own models had similarly gained unauthorized access to the live systems of multiple organizations that had not detected the intrusions. Jacob Coxon, a researcher who had worked at both companies, recently resigned from Anthropic and warned that OpenAI and Anthropic are “gambling with our lives.”

To be clear, I am not arguing against the technology. I use AI tools nearly every day, including in preparing this essay, and I think they have enormous potential to benefit our world. But the dangers cannot be ignored. The question is how we as a society can harness the benefits while taking the risks seriously. We need a way to responsibly move forward with a technology that is simultaneously capable of curing a disease or building a bioweapon, of shoring up our cyber defenses or taking down a power grid, of helping us address climate change or contributing to the problem, and of enhancing human productivity or replacing a significant percentage of human labor.

AI engineers are positioned well to understand the dangers of what they have built, but decisions about whether and how those risks should be regulated, and who will bear the costs, should be made with all stakeholders at the table, including the legal profession. Many lawyers are already involved in this work, such as through the American Law Institute’s Principles of the Law, Civil Liability for Artificial Intelligence, but more of us can be helpful across an unusually wide array of legal fields.

Some connections are clear. For example, as the ALI project illustrates, tort lawyers and scholars can help develop rules for allocating responsibility among model developers, deploying entities and users when foreseeable harms occur. Take the recent cybersecurity incidents: What duties should govern testing, who should be notified when safeguards fail, how should responsibility be shared with the third parties conducting the evaluations, and what remedies should we give to those who are adversely affected?

Other fields of law are just as relevant but perhaps less obviously so. Administrative law specialists, for instance, can assess whether existing agencies have the authority and expertise to oversee AI. If a new agency is warranted, what powers should Congress grant, how should its decisions be reviewed, and how can it resist capture by the companies it regulates?

Tax policy is one possible tool for addressing concerns about the potential for widespread job losses. One recent paper proposes a tax on AI tokens (the units used to meter many AI services) that would, its proponents argue, remove the tax code’s tilt toward automation and protect tax revenue as labor income falls. Working through ideas like this one, including the effects on innovation and affordability, is the kind of analysis lawyers are trained to do.

Environmental and energy lawyers are already grappling with where to build data centers and how to manage their heavy demands for electricity and water. They are working on a wide range of issues, including how permitting can support useful infrastructure while protecting communities, who should bear the costs of expanding the grid, and how other customers can be protected from higher rates.

The implications for human rights, civil rights and international law are also significant. Among other issues, experts in those fields can help define limits on the government’s use of AI for surveillance and autonomous weapons, address algorithmic discrimination, and develop appropriate responses to the proliferation of AI-generated misinformation.

Lawyers also have important insights to share about the unintended consequences of overregulation. For example, what if new regulations slow down model development and make the U.S. more susceptible to cyberattacks from unfriendly countries or terrorist organizations? Or enable those countries to gain economic advantages that adversely affect American industries?

Many lawyers have helped clients navigate well-intentioned regulations that have had adverse effects, including regulations of the legal profession itself. Proposals to license access to high-risk AI capabilities deserve the same scrutiny because a poorly designed regime could entrench incumbents while doing little for safety.

Of course, we can’t and shouldn’t do this work alone; engineers, economists, policymakers and affected communities all should play a role. But we need people who can translate technical warnings into workable protections and clear allocations of responsibility, and we also need people to warn when regulations might cause more harm than they are worth. Lawyers can and should help to do this work.

Here’s a thought exercise. Look up a pending bill, proposed rule or model-law draft that relates to your field. Congress is weighing several bills right now, and many states are moving even faster. What does your own practice reveal that the drafters may have overlooked? I suspect most lawyers who try this exercise will find they have something helpful to say. If you do, send your thoughts to the bill’s sponsor, a relevant agency or a bar committee.

Choices about public policy should be made through accountable public institutions, and the legal profession can play a valuable role in ensuring that those choices are informed by lawyers’ distinctive perspectives. Our responsibility, in short, is not only to learn how to use AI to help our clients. It is to help the public govern it.

Andrew M. Perlman is the dean and a professor at the Suffolk University Law School and the vice president for academic innovation at Suffolk University. He has worked as a member of the advisory council of the ABA Task Force on Law and Artificial Intelligence, as the chief reporter for the ABA Commission on Ethics 20/20, as vice-chair of the ABA Commission on the Future of Legal Services, and as the inaugural chair of the governing council of the ABA Center for Innovation. He writes in his personal capacity. In preparing this essay, he used OpenAI’s GPT-6 Astra and Anthropic’s Claude Fable and Opus models for background research and for suggestions on structure and phrasing; he takes full responsibility for the essay’s content, accuracy and attributions.

ABAJournal.com is accepting queries for original, thoughtful, nonpromotional articles and commentary by unpaid contributors to run in the Your Voice section. Details and submission guidelines are posted at “Your Submissions, Your Voice.”