As artificial intelligence (AI) tools become increasingly integrated into hiring and other workplace decisions, businesses must navigate a rapidly evolving legal landscape regulating the use of AI. To stay compliant and build trust within the workforce, employers can consider the following best practices for responsible AI deployment in employment contexts.
- Maintain human oversight. Use AI as a tool to assist, not replace, human judgment. Require managers to confirm AI outputs before relying on them to make employment decisions, including for hiring, promotions, work allocation, compensation, and discipline.
- Promote transparency. Notify applicants and employees when AI tools are being used to evaluate them. Consider explaining the extent to which managers or human resources (HR) will review the AI tool’s outputs, how the AI works, how it will impact decision-making, and the data the AI tool uses.
- Review for potential bias. Periodically audit the outputs of AI tools for potential bias. For example, compare AI decisions with human decisions, explore whether similar candidates—except for a protected attribute like gender, race, or ethnicity—are receiving different outcomes, and conduct disparate impact analyses. Document any bias testing for use in defending against potential discrimination claims.
- Train key personnel. Provide training to HR professionals, hiring managers, and others involved in deploying or interpreting AI outputs, ensuring they understand the tool’s capabilities, limitations, and compliance obligations. Training can include topics such as how the AI works, recognizing when AI outputs are inaccurate or incomplete, and how to override or edit AI-generated content. Encourage employees to escalate any issues or inaccuracies in AI outputs.
- Maintain data privacy and security. Implement safeguards to protect the data fed into any AI tool, particularly any sensitive or personal employee information.
- Ensure records retention. Retain records regarding AI use in employment decisions, in accordance with data retention policies or as required by applicable laws.
- Screen AI vendors. Carefully vet third-party AI providers to ensure their tools meet legal standards and provide relevant documentation, such as to confirm steps taken to validate the tool and mitigate bias.
- Monitor legal developments. As the AI regulatory environment is rapidly evolving, stay up to date on federal, state, and local laws that regulate or impact the use of AI in the workplace.
If you have any questions concerning our practices, please contact the following members of our firm: Lindsay Burke, Carolyn Rashby, and Michelle Barineau York.