AI Governance Gaps Exposed: Most Firms Lack Effective Systems, AAA Survey Finds — American Arbitration Association

Event summary

87% of large U.S. and Canadian firms have AI governance frameworks, but only 22% say they work effectively.
33% of respondents lack defined escalation pathways for AI system misbehavior.
22% are confident in their ability to produce evidence of governance decisions for regulators.
Only 21% of CEOs hold final AI deployment authority, suggesting delegation of governance responsibilities.

The big picture

The survey highlights a critical disconnect between AI governance policies and their real-world effectiveness, as firms struggle to manage AI risks amid expanding deployments. With 70% of respondents representing billion-dollar organizations, the findings underscore the cross-functional imperative of AI governance across industries. Effective governance requires practical systems that integrate oversight, accountability, and decision-making throughout the AI lifecycle, particularly as regulatory expectations evolve.

What we're watching

Governance Dynamics: How firms will bridge the gap between AI governance policies and practical implementation.
Regulatory Headwinds: Whether limited audit readiness will expose firms to increased regulatory scrutiny.
Leadership Involvement: The pace at which CEOs will take direct ownership of AI deployment decisions.