When corporations use Artificial Intelligence (AI) programs, they should be sure to use them in a manner that doesn’t result in the Department of Justice knocking on their door. The DOJ just released an updated version of its "Evaluation of Corporate Compliance Programs," a document used by the DOJ when deciding whether to conduct an investigation, whether to bring charges, and while negotiating plea agreements. This revision is notable because it includes significant new content directly addressing the use of artificial intelligence in corporate operations and compliance programs.
Here are some key takeaways:
- Risk Assessment: Companies must assess and address risks associated with AI use in their operations. The DOJ will evaluate whether companies have done so and taken appropriate steps to mitigate risks associated with AI use.
- Policies and Procedures: Companies should update policies and procedures to address emerging risks, including those associated with new technologies like AI.
- Training and Communications: Employee training should cover lessons learned from compliance issues faced by other companies in the same industry, which may include AI-related incidents.
- Monitoring and Testing: Companies using AI in commercial operations or compliance programs must:
- Monitor and test AI technologies to ensure they function as intended and align with the company's code of conduct.
- Implement systems to quickly detect and correct AI decisions that are inconsistent with company values.
- Continuous Improvement: The guidance emphasizes the need for companies to:
- Review and adapt compliance programs based on lessons learned from their own misconduct and that of other companies facing similar risks.
- Ensure the ability to evaluate whether AI and other new technologies are functioning as intended and consistent with the company's code of conduct.
- AI Governance: The DOJ will assess a company's approach to AI governance, including:
- Integration of AI risk management into broader enterprise risk management strategies.
- Measures to curb potential negative or unintended consequences of AI use.
- Controls to monitor and ensure AI trustworthiness, reliability, and compliance with applicable laws and company codes of conduct.
- Accountability measures for AI use.
To avoid potential DOJ penalties or investigations, companies using AI should proactively address these areas in their compliance programs. Companies should review their current compliance programs in light of this updated guidance and make necessary adjustments to address AI-related risks and governance issues. An ounce of prevention is better than a pound of cure.
