Biden signs executive order on rules for generative AI, requiring federal agencies to develop safety guidelines

US President Joe Biden has signed an executive order setting rules and safety standards for the field of generative artificial intelligence, ahead of lawmakers.

Here's What We Know

The executive order has eight goals:

• Developing new AI safety and security standards;
• protecting privacy;
• promoting equality and civil rights;
• protecting consumers, patients, and students;
• supporting workers;
• promoting innovation and competition;
• promoting U.S. leadership in AI technologies;
• ensuring responsible and effective government use of technology.

The White House has directed a number of agencies to develop cybersecurity standards, prevent the use of AI to create bioweapons, and form programmes to prevent discrimination and ensure fairness in AI-based decision-making.

Large companies developing AI models like OpenAI's GPT and Meta's Llama will be required to share the results of security testing of their systems.

A White House spokesperson noted that most of the requirements apply to new AI models. The executive order does not apply to existing and publicly available systems.

The executive order is effective immediately, but is not permanent law. It is only in effect under this administration. In the meantime, Congress continues to debate legislative regulation of the AI field.

Source: The Verge