Business · Companies · 1 hr ago
Zuckerberg says Meta’s new AI model puts it at the frontier
Meta CEO Mark Zuckerberg says the company’s new Muse Wonder model puts it at the forefront of artificial intelligence.
Muse is an app that uses AI agents, and it debuted in early September.
Zuckerberg said Meta spent heavily on computing power and years of scientific work to develop the model.
Meta is also planning to sell access to its latest Muse models to businesses, putting it in competition with Anthropic and OpenAI.
The app’s growth has drawn interest from analysts, while businesses such as travel agencies and tax preparers are concerned that its capabilities could threaten their services.
The expansion also carries risks: Meta faces questions about the cost of building data centers, security flaws reportedly found before launch, and the personal data its products collect.
Zuckerberg says he does not support slowing AI development, and argues that AI agents still need better common sense to become reliable personal assistants.
Meta CEO Mark Zuckerberg says the company’s new Muse Wonder AI model puts it at the frontier of AI development.
He told the Financial Times that the model could support more sophisticated behaviors and said its development required substantial computing power, investment and research.
Zuckerberg said he does not support slowing AI development, unlike a proposal by Anthropic CEO Dario Amodei that OpenAI CEO Sam Altman endorsed.
He said Meta is focused on personal empowerment and argued that AI agents still lack the common sense needed to be ideal personal assistants.
Muse’s growth and agentic abilities have drawn interest from analysts and concern from sectors that could face competition.
The article also raises questions about the cost of Meta’s AI infrastructure and the model’s security and use of personal data.
- Who
- Meta CEO Mark Zuckerberg, speaking about the company’s Muse Wonder AI model.
- What
- Zuckerberg said the model puts Meta at the AI frontier. He also said he does not support slowing AI development.
- When
- Monday; the article was published on October 13, 2026.
- Where
- Not stated
- Why
- Zuckerberg said the model’s development required substantial computing power, investment and scientific work, and said AI agents still lack enough common sense to be ideal personal assistants.
Mark Zuckerberg
Dario Amodei and Sam Altman
Slowing AI development
Mark Zuckerberg
Zuckerberg said he does not support slowing AI development.
Dario Amodei and Sam Altman
Amodei proposed slowing AI development, and Altman endorsed the proposal.
Meta’s Muse app debuted.
Zuckerberg told the Financial Times that Muse Wonder could support more sophisticated AI behaviors and said he opposed slowing AI development.
Gizmodo published its article about Zuckerberg’s comments.
- New model
- Muse Wonder
- Company
- Meta
- CEO
- Mark Zuckerberg
- Muse debut
- Early September
- Muse traffic ranking
- No. 1 on Human Security’s list of top agent-based traffic sources, accounting for 72% of the total, according to Barron’s
- Projected infrastructure spending
- At least $600 billion by the end of 2028, as projected by Meta as of October 2025
What people are saying
allow much more sophisticated behaviors that are at the frontier of what is being done in the AI field.
Getting to this point both required a large amount of compute and investment, and a lot of very rigorous scientific work by the group working on it over an extended period of time.
Everyone else is focused on more knowledge work, automation-type cases, and we’re much more focused on personal empowerment
Most of the ways in which people might not trust the company can lead to technological innovations that could be quite productive






