Dual-Board Model Would Make AI Safety a Fiduciary Duty, Brooklyn Law Review Article Argues
What happened
A peer-reviewed article titled Governing Artificial Intelligence: A Dual-Board Solution, published in the Brooklyn Law Review, proposes a structural reform to U.S. corporate governance that would require companies deploying consequential AI systems to establish a second, specialized board with explicit fiduciary responsibility for AI safety oversight. The proposed AI safety board would operate alongside the traditional board of directors, holding distinct legal duties focused on identifying, monitoring, and mitigating AI-related harms. The article frames AI safety not as a discretionary management concern but as an enforceable obligation potentially actionable through shareholder litigation and regulatory scrutiny. The authors draw an explicit parallel to the audit committee model mandated by Sarbanes-Oxley, arguing that AI safety demands comparable structural independence and legal accountability. The article also engages with whether voluntary frameworks such as the NIST AI Risk Management Framework and ISO 42001 provide sufficient enforcement mechanisms at the corporate level.
Why it matters
- ·The proposed dual-board model signals a legal theory under which companies with high-stakes AI deployments could face shareholder litigation or regulatory action if their board committee structures are deemed inadequate for AI safety oversight, creating new dimensions of regulatory exposure particularly under SEC reporting requirements and state statutes such as Colorado's SB 26-189, which replaced the repealed Colorado AI Act.
- ·Operationally, the article implies that existing AI governance documentation including risk registers, safety testing records, and incident response protocols may need to demonstrate board-level engagement to be considered legally adequate, raising the evidentiary bar for enterprise AI compliance programs.
- ·Organizations whose boards rely on single-board structures or informal AI subcommittees face reputational and fiduciary risk if institutional shareholders or plaintiffs' attorneys adopt the dual-board standard as a benchmark for evaluating whether leadership took AI risk seriously following an AI-related incident.
Governance controls affected
What to do now
- ☐Conduct a gap assessment of current board committee structures against the dual-board accountability model described in the article, with particular focus on sectors such as financial services, healthcare, and critical infrastructure.
- ☐Review existing AI risk registers and safety testing records to determine whether they document board-level engagement sufficient to satisfy an evolving fiduciary duty standard.
- ☐Brief general counsel and the chief compliance officer on the legal theories in the article so they can evaluate exposure under SEC AI disclosure guidance and Colorado's SB 26-189, which replaced the repealed Colorado AI Act.
- ☐Verify that incident response playbooks and post-incident review processes include escalation paths that reach board-level oversight, not only management-level functions.
- ☐Monitor institutional shareholder communications and proxy advisor guidance for emerging references to dual-board or equivalent AI safety oversight standards that could influence upcoming annual meeting resolutions.
What to watch next
Compliance teams should monitor whether the legal theories advanced in this article are cited in SEC comment letters, shareholder proposals, or state regulatory guidance as a benchmark for adequate AI board oversight. The evolution of the Colorado AI Act and any analogous state-level statutes may incorporate board-level accountability language influenced by scholarly frameworks such as this one. Teams operating in heavily regulated sectors should also track whether plaintiffs' firms begin referencing the dual-board standard in securities or derivative litigation following AI-related incidents.
Stay ahead of stories like this
Get every US AI governance development like this one, plus the rest of the week's developments. Every Thursday.
Recent issues
- Regulators, litigants, and attackers are all moving faster than most organizations' AI governance frameworks, shifting leverage away from deployers and toward those who can prove control, accountability, and honest disclosure.21 Oct
- Three separate incidents this week reveal that AI vendors cannot be trusted to self-police, and that harm from deployed agents can run unchecked for months before anyone inside the organization notices.15 Oct
Free every Thursday. Unsubscribe anytime.