President Joe Biden has signed an executive order outlining rules and guidelines for generative AI. This order comes as a preemptive measure, ahead of any legislative action by lawmakers, and carries eight primary objectives.
The executive order aims to: create new standards for AI safety and security, protect user privacy and data, advance equity and civil rights in AI applications, advocate for consumers, patients, and students in AI-related matters, support workers impacted by AI technologies, promote innovation and competition in the AI sector, ensure the United States maintains its leadership in AI technologies, and ensure responsible and effective government use of AI.
The order assigns various government agencies specific tasks to achieve these goals. Notably, the National Institute of Standards and Safety (NIST) will be responsible for developing standards to rigorously test AI models before public release. Additionally, the Department of Energy and the Department of Homeland Security will address the potential threat of AI to infrastructure, including cybersecurity risks.
Developers of large AI models, such as OpenAI’s GPT and Meta’s Llama 2, are required to share safety test results. Importantly, the safety guidelines primarily apply to future AI models and will not recall publicly available models that already exist, while anti-discrimination rules remain in place.
The order also focuses on privacy, urging Congress to pass data privacy regulations and promoting the development of privacy-preserving AI techniques.
Furthermore, it seeks to prevent AI discrimination, especially in sentencing, parole, and surveillance. Government agencies are instructed to provide guidelines to prevent AI from exacerbating discrimination in various sectors, including housing, federal benefit programs, and contracts.
The order also addresses the potential impact of AI on the labor market, calling for a report on job displacement. To support AI research and development, the White House plans to establish a National AI Research Resource, providing key information to students and researchers and offering technical assistance to small businesses. It also encourages the rapid hiring of AI professionals for the government.
While this executive order is a significant step towards regulating AI, it is not permanent law and is limited to the duration of President Biden’s administration. Lawmakers are currently discussing more comprehensive legislation regarding AI.
Industry observers, including Navrina Singh, founder of Credo AI and a member of the National Artificial Intelligence Advisory Committee, consider this executive order a crucial step towards demonstrating the seriousness of the United States about generative AI. It sets the tone for AI policy and regulation, even as more comprehensive laws are under discussion.