Notes:

  • Emerging Risk Landscape: AI introduces unique risks: algorithmic bias, hallucinations, data privacy violations, intellectual property infringement, and safety failures (e.g., autonomous vehicle accidents).
  • California’s TFAIA (Transparency in Frontier Artificial Intelligence Act, 2026):
    • Requires “large frontier developers” to adopt and publish a Frontier AI Framework.
    • Governance Requirements: Must define thresholds for catastrophic risk, apply mitigations, and use third-party assessments.
    • Reporting: Must report critical safety incidents to the state within 24 hours if there is imminent risk of death/injury.
    • Whistleblower Protection: Specific protections for employees reporting AI risks.
  • Board Responsibilities:
    • Risk Assessment: The Board must oversee the identification of AI risks specific to the company’s use cases (e.g., hiring algorithms, customer service chatbots, trading bots).
    • Ethical Guidelines: Ensuring the company has an AI Ethics Policy that addresses bias, fairness, and transparency.
    • Data Governance: Overseeing the quality, security, and privacy of the data used to train and operate AI models.
  • Regulatory Compliance: Boards must stay abreast of evolving AI regulations (EU AI Act, US Executive Orders, state laws) and ensure the company’s AI governance framework is compliant.
PreviousNext