Archive · 16 Sep 2026
AI Companies Debate Safety Slowdown Amid Internal Tensions
Some companies that build advanced AI are trying to make the technology safer.
Anthropic’s leader, Dario Amodei, warned that AI is becoming more powerful and said its development may need to slow down.
OpenAI’s leader, Sam Altman, has also supported safety steps, including independent experts checking AI systems.
OpenAI, Anthropic and Google DeepMind have been talking about these issues for several weeks.
They are considering shared safety rules and an industry group to set standards.
Anthropic suggested that the government could give companies limited permission to cooperate without antitrust problems.
OpenAI policy chief Chris Lehane said such permission is not necessary for the companies to work together.
The effort has reportedly caused disagreements inside the companies about security and competition.
President Donald Trump disagrees with slowing AI development and calls the safety concerns a “hoax.”
Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman have urged stronger safeguards as AI systems become more capable.
OpenAI, Anthropic and Google DeepMind have coordinated on AI safety discussions for several weeks.
The companies are discussing independent evaluators, an industry standards body and possible industry-wide responses to AI risks.
Anthropic proposed a limited government waiver for safety coordination, while OpenAI policy chief Chris Lehane said the companies do not need one.
The coordination has reportedly created internal tensions, while US President Donald Trump has called AI safety concerns a “hoax” and opposed development limits.
- Who
- Anthropic, OpenAI, Google DeepMind and their leaders and policy officials, including Dario Amodei, Sam Altman and Chris Lehane; President Donald Trump opposes a slowdown.
- What
- The companies are coordinating on AI safety, possible development slowdowns, independent evaluations and industry standards, amid reported internal tensions.
- When
- The companies have been communicating for several weeks, following Amodei’s 3,800-word essay on AI risks.
- Where
- The articles do not specify a location; the discussions involve major AI companies and the US government.
- Why
- Supporters seek safeguards as AI capabilities grow, while Trump opposes guardrails and restrictions on AI development.
Supporters of AI Safeguards
Opponents of an AI Slowdown
Development pace
Supporters of AI Safeguards
Dario Amodei and other safety advocates argue that increasingly capable AI may require slowing development and adding stronger safeguards.
Opponents of an AI Slowdown
Donald Trump has rejected calls to slow AI development and has criticized efforts to impose limits on advanced AI systems.
Oversight and guardrails
Supporters of AI Safeguards
Supporters favor government oversight, regulation, independent evaluators and an industry standards body to assess AI risks.
Opponents of an AI Slowdown
Trump has dismissed the safety concerns as a “hoax” and opposed the need for guardrails or restrictions.
Industry coordination
Supporters of AI Safeguards
OpenAI, Anthropic and Google DeepMind are coordinating on safety and discussing shared standards; supporters view cooperation as a way to address common risks.
Opponents of an AI Slowdown
The coordination has raised concerns about antitrust law, competition and internal security, with some leaders warning that companies must avoid suppressing competition.
- Companies involved
- OpenAI, Anthropic and Google DeepMind
- Key warning
- Anthropic CEO Dario Amodei published a 3,800-word essay warning about growing AI capabilities.
- Safety measures
- Independent third-party evaluators, government oversight, regulation and industry standards
- Coordination period
- The companies have been discussing AI safety for several weeks.
- Waiver proposal
- Anthropic proposed a limited government waiver to permit safety coordination without triggering antitrust concerns.
- OpenAI position
- Chris Lehane said the companies do not need a waiver to work together on safety.
- Opposition
- Donald Trump called AI safety concerns a “hoax” and criticized guardrails or restrictions on development.
