The White House Office of Management and Budget (OMB) has released a memo outlining new policies for the use of AI in the US government, aiming to address concerns about transparency, accountability, and fairness.
In recent years, the world has been captivated by the advancements and controversies surrounding AI research. However, amidst the buzz surrounding high-profile events and breakthroughs, the US government has quietly been using AI systems for more mundane tasks. From tracking undocumented migrants to allocating police resources, AI has become a common tool for cutting costs. Yet, this reliance on AI has raised concerns about the potential for arbitrary rule and discriminatory outcomes. Finally, the call for transparency and accountability in the use of AI in government is being answered with the release of a groundbreaking memo by the OMB.
A Revolution in Transparency and Accountability
The OMB, the largest unit in the US president’s executive office, has released a draft policy memo that could revolutionize the use of AI in government. The proposed policy requires each department to appoint a chief AI officer and provide a register of existing AI use cases, a significant win for transparency. Additionally, the memo mandates the identification of AI systems that may impact people’s safety or rights, subjecting them to meaningful constraints.
Weighing Risks and Benefits
Under the new policy, agencies must carefully weigh the risks to rights and safety posed by AI systems against their claimed benefits. This requirement aims to address the inherent biases in AI models and the unrepresentative data on which they are often trained. Government departments must also consult affected groups before deploying AI systems and provide options for human consideration and remedy, allowing individuals to contest decisions that adversely affect them.
Addressing Government Procurement of AI
The OMB memo goes beyond transparency and accountability by addressing the important question of government procurement of AI. It highlights the need for best practices in the procurement process, aiming to prevent inexperienced government agencies from adopting software they do not fully understand. By requiring government departments to follow these best practices, the memo aims to mitigate the societal problems that arise from the failure of AI systems, particularly affecting the most vulnerable.
A Model for Research and Civil Society-led Policymaking
The OMB memo serves as a case study in research and civil society-led policymaking. In contrast, attempts to regulate frontier AI models in the EU have faced challenges. The EU parliament’s attempt to regulate AI systems, including ChatGPT, highlighted the difficulties of regulating a rapidly evolving field. The lack of time for public interest research and civil society debate has led to disarray in the regulatory process. The OMB memo provides a more balanced and well-grounded approach to AI governance.
Conclusion:
The OMB memo represents a significant step towards responsible AI governance in the US government. By prioritizing transparency, accountability, and fairness, the memo addresses the concerns surrounding the use of AI in government systems. It sets a precedent for other countries and organizations to follow in their efforts to regulate AI. However, the challenges of regulating frontier AI models highlight the need for further research and civil society debate to inform more mature policies. As AI continues to advance, it is crucial to strike a balance between innovation and responsible governance to ensure the equitable and ethical use of AI in society.
