FBR Allows Instalment Payments for Imported Phone Taxes
September 15, 2026
Anthropic CEO Dario Amodei has urged AI companies to slow the pace of model development, proposing stronger independent oversight and industry coordination as concerns grow over the misuse and safety risks of advanced AI systems.
Anthropic CEO Dario Amodei has called on artificial intelligence companies to moderate the pace at which they develop increasingly capable AI models, arguing that developers need more time to address growing safety and misuse risks.
In an essay published as concerns around advanced AI intensify, Amodei proposed a three-step framework focused on slowing capability improvements, strengthening independent oversight and encouraging greater cooperation across the AI industry.
According to Business Recorder’s report, Amodei said, “We must slow the pace at which we improve the capabilities of AI models,” while stressing that progress would remain fast and that companies should make effective use of the additional time.
His comments followed an Anthropic threat intelligence report detailing how Claude AI models had been used in activities including weapons development, cyber operations, surveillance and fraud. The incidents have added to concerns over whether increasingly capable AI systems can be adequately controlled once deployed.
Amodei clarified that he was not calling for an end to AI training or technical progress. Instead, he argued that companies should give greater attention to alignment, safety testing and independent evaluation before rapidly advancing model capabilities.
One of his proposals is for permanent third-party reviewers to operate inside frontier AI companies, with access to relevant tools and internal risk-assessment processes. The aim would be to provide independent verification that appropriate safeguards are being implemented.
TechCrunch reported that Amodei’s proposal comes amid increasingly public warnings from AI researchers and executives about the potential consequences of rapid progress. The discussion has increasingly centred on whether the industry needs to “pace the frontier” rather than pursue capability improvements at maximum speed.
The debate has also been fuelled by incidents involving AI agents accessing external systems and operating beyond their intended environments. Recent disclosures involving Anthropic and OpenAI have intensified scrutiny over the ability of developers to contain increasingly autonomous systems.
Amodei also called for AI companies to work together voluntarily to establish common safety standards, while encouraging international cooperation on managing advanced AI risks.
The proposal comes as governments and lawmakers in the United States and elsewhere consider stronger rules for artificial intelligence. However, industry-wide coordination could prove difficult because companies remain in intense competition to develop more capable systems.
For Anthropic, the message is not to stop AI innovation but to create enough time for safety measures, independent evaluation and governance mechanisms to keep pace with rapidly advancing capabilities.
