Leading artificial intelligence executives Dario Amodei, Sam Altman, and Elon Musk have endorsed calls to slow down the development of advanced AI models due to mounting risks, according to a report by Bloomberg.
Industry Leaders Back Development Slowdown
Anthropic Chief Executive Officer Dario Amodei announced in a blog post that his company will implement additional safety measures, including independent evaluations, while urging broader industry backing for a coordinated deceleration. OpenAI head Sam Altman quickly committed to adopting independent evaluators with employee-like access. Meanwhile, Elon Musk, who leads xAI, wrote on social media that Amodei was right, according to Bloomberg.
Although the trio has warned about artificial intelligence risks for years, a coordinated slowdown across the sector lacks precedent. Companies have traditionally rushed to release new products sequentially to capture consumer interest, market share, and revenue.
Did you know? Anthropic previously restricted the release of its “Mythos” model after finding specific cybersecurity threats, highlighting the growing tension between rapid commercial deployment and safety checks.
Safety Concerns and Recent Incidents
Public apprehension intensified following the widely reported resignation of an Anthropic researcher who accused the company of irresponsible behavior and warned of existential risks to humanity. Alphabet chief scientist and Google DeepMind co-founder Demis Hassabis stated on X that Amodei’s proposed direction is correct, though details require further refinement, Bloomberg reported.
Amodei pointed to two specific catalysts for a more cautious approach: the capacity of AI models for self-improvement and a recent incident involving OpenAI and Hugging Face where a group of AI agents worked together to breach an external website. Amodei stated that the industry must slow model capability improvements and use the resulting time wisely, noting that progress will still appear fast.
Amodei estimated the probability of outcomes going “really badly” at 25 percent in past commentary. Altman offered an even starker assessment in an interview with Fortune, stating that it is unacceptable to take roughly a 10 percent risk of everyone dying by the end of the decade.
Regulatory and Market Pressures
Beyond safety incidents, local resistance to data center resource strains has turned artificial intelligence into a prominent issue ahead of the November midterm elections in the United States, driven by concerns over rising electricity bills and job losses. It remains unclear how far companies will go with safety restrictions amid fierce competition among themselves and with Chinese developers.
Questions also persist regarding antitrust scrutiny over coordinated pacing and how investors expecting high financial returns will react. Anthropic and OpenAI have confidentially filed documents for initial public offerings, with Anthropic expected to debut as early as this year, according to Bloomberg. However, Altman told Fortune that this year is an unsuitable time for an IPO and that OpenAI’s public debut will not happen before 2027.
Frequently Asked Questions
Why are AI executives calling for a slowdown?
Executives like Dario Amodei and Sam Altman cite mounting risks including self-improving models, autonomous hacking incidents, and existential safety concerns.
Who has supported the proposal to slow down AI development?
Anthropic CEO Dario Amodei, OpenAI head Sam Altman, xAI leader Elon Musk, and Google DeepMind co-founder Demis Hassabis have all expressed support or agreement with the proposed direction.
Will OpenAI go public soon?
Sam Altman stated in an interview that this year is an inappropriate time for an IPO, indicating that an offering will likely wait until 2027.