Commerce Unveils New AI Safety Guidance

The U.S. Department of Commerce marked a milestone 270 days after President Biden's Executive Order on AI by releasing new guidance and tools to enhance AI safety, security, and trustworthiness.

NIST unveiled three final guidance documents and a draft from the U.S. AI Safety Institute, addressing risks from generative AI and dual-use foundation models. Alongside this, a software package was introduced to assess the impact of adversarial attacks on AI performance.

The USPTO updated patent guidelines to clarify AI innovation eligibility, while the NTIA submitted a report on the risks and benefits of large AI models with open weights.

Commerce Secretary Gina Raimondo emphasized the urgency of developing safe AI systems, highlighting the ongoing efforts to keep America at the forefront of AI innovation.

AI's New Rulebook:

NIST's draft guidance aims to help developers mitigate the risks of generative AI and dual-use models.
New software measures AI systems' resilience against adversarial attacks.
USPTO's updated patent guidelines provide clarity on AI-related inventions.
NTIA's upcoming report focuses on the balance of risks and benefits of open-weight AI models.

Why it matters: These initiatives reflect the government's proactive stance on AI safety, crucial for maintaining global leadership in technology while ensuring innovation does not compromise security.

0 comments