Tech Giants Forge Ahead With AI Safety Standards Amid Criticism

WHO:

Techmeme - Google, OpenAI, and Anthropic are involved in creating an AI safety standards body.

NBC News - Mark Zuckerberg is the founder and CEO of Meta.

NBC News - Dario Amodei is the CEO of Anthropic, and Sam Altman is the CEO of OpenAI.

NBC News - Jacob Coxon is a former researcher at Anthropic.

NBC News - Demis Hassabis is a co-founder of Google DeepMind, and Elon Musk is associated with xAI.

WHAT:

Techmeme - Google, OpenAI, and Anthropic plan to launch an AI safety standards body without government oversight.

NBC News - Zuckerberg rejected calls for an industrywide slowdown of AI development.

NBC News - Zuckerberg stated that Meta's Muse AI agent has become the most popular free offering on the Apple App and Google Play stores.

NBC News - Muse AI can perform various tasks, including booking plane tickets and replying to emails.

NBC News - Zuckerberg mentioned that Muse is composed of several agents, including Sentinel, which oversees the main Muse agent.

WHERE:

NBC News - The interview with Zuckerberg took place in California.

WHEN:

Techmeme - The AI safety standards body is planned to launch in late 2026 or 2027.

NBC News - The Muse AI agent was released on September 8.

NBC News - Amodei and Altman briefed the United Nations Security Council on a Wednesday.

WHY:

NBC News - Zuckerberg believes that each lab should internally ensure safety rather than coordinate industrywide efforts.

NBC News - Concerns about AI systems spiraling out of human control prompted calls for a slowdown in development.

HOW:

NBC News - Zuckerberg emphasized that AI companies should focus on aligning their systems with users' desires and handling information with discretion.

NBC News - Zuckerberg stated that Muse and other AI agents have privacy concerns due to the amount of information users provide.

(UPDATED September 24, 2026 14:02 UTC)

Google, OpenAI, and Anthropic are set to launch an AI safety standards body without government oversight, aiming for a rollout in late 2026 or 2027. In a recent interview in California, Meta's CEO Mark Zuckerberg rejected calls for an industrywide slowdown in AI development, highlighting that Meta's Muse AI agent has quickly become the most popular free offering on app stores and can perform tasks like booking plane tickets and replying to emails. Zuckerberg also noted that Muse is made up of several agents, including Sentinel, and emphasized the importance of AI companies ensuring their systems align with user desires while addressing privacy concerns.