Policymakers and companies alike are working to ensure the responsible innovation, development, deployment, and adoption of artificial intelligence (AI). Comprehensive corporate governance initiatives for AI are critical to further enhancing transparency and conveying the products’ safety and reliability to customers.
Companies with insight into the AI value chain are uniquely positioned to lay the groundwork for policy by enacting proactive and thoughtful organizational safeguards against risk. Every company should pursue some type of governance for their AI adoption, and the appropriate governance mechanisms should correspond to their role within the value chain and to how they are developing and deploying AI.
BSA’s “AI Policy Solutions” document details how companies can develop their corporate governance practices to mitigate risks and biases in AI systems and technologies. Our latest document, “Best Practices for AI Governance,” elaborates on these recommendations with steps that companies can take to mitigate the risk of bias throughout the entire AI lifecycle.
BSA’s Best Practices Include:
- Emphasizing AI leadership, oversight, and accountability;
- Investing in trained personnel and sufficient resources for AI;
- Collaborating with partners to ensure responsible AI use by all stakeholders;
- Enhancing transparency with clear communication and necessary documentation; and
- Establishing risk mitigation strategies, including impact assessments.
Why It Matters:
Best practices for AI governance should focus on mitigating the risk of bias throughout the AI lifecycle by implementing risk mitigation procedures such as conducting AI impact assessments for the organization. This not only helps to protect customers but also ensures a company’s use of AI is responsible and in line with best practices.
What’s Going on:
- In February, BSA joined the US AI Safety Institute Consortium, which brings together more than 200 stakeholders to support the safe development and deployment of AI.
- Technology companies signed Seoul’s AI Business Pledge during the AI Summit in May. BSA issued a statement supporting the BSA members who signed.
- President Biden issued his executive order on AI last fall, pulling from the National Institute of Standards and Technology’s AI Risk Management Framework as a resource.