A proposed Big Tech AI slowdown has ignited debate over whether leading developers are prioritising safety or protecting their market dominance. Over a recent weekend, several of the most prominent figures in artificial intelligence loosely agreed to slow the pace of AI development, a move that sceptics quickly viewed with suspicion.
OpenAI chief executive Sam Altman, Anthropic chief executive Dario Amodei, Google DeepMind cofounder Demis Hassabis, and SpaceX head Elon Musk were among those who signed on, at least partially, to the initiative. Their stated aim was to “pace the frontier” of AI progress, framing the agreement as a measure intended to manage risks tied to rapid technological advancement.
What the AI leaders proposed
The proposal set out several concrete steps. It called for embedding third-party auditors to review the work of AI developers, regulating domestic laboratories, and pursuing a global slowdown agreement across the industry. Together, these measures were presented as a framework for more cautious and accountable development of advanced systems.
Supporters of the plan argue that coordinated restraint is necessary given the speed at which frontier models are being built and deployed. By introducing external oversight and shared limits, the signatories suggest the industry could reduce the potential for harm while still advancing the technology in a controlled manner.
Why critics see a cartel
Not everyone accepts the safety rationale. Critics contend that the agreement carries an ulterior motive, arguing that the companies involved may be seeking to block potential rivals from entering the market. Some maintain that the plan could weaken the open-source movement, which relies on the free and rapid sharing of models and research.
Others go further, suggesting that a coordinated slowdown among the most powerful players amounts to anti-competitive behaviour. Detractors have raised concerns that the arrangement allows the firms to avoid meaningful legal safeguards while consolidating their positions. Some observers have described the pact in the strongest terms, labelling it an outright cartel.
The disagreement highlights the tension at the heart of the current AI landscape. On one side sits the argument that established developers are best placed to set responsible limits and invite scrutiny through independent audits. On the other is the claim that such limits conveniently serve the commercial interests of the companies proposing them, by curbing competition and discouraging alternative approaches to building AI.
The initiative brings together individuals who lead or helped found some of the most influential organisations in the sector, including OpenAI, Anthropic, Google DeepMind, and SpaceX. Their willingness to align, even loosely, on a shared framework signals a notable moment of coordination among firms that are otherwise direct competitors in the race to develop advanced artificial intelligence.
Source
Image: theverge.com
