Anthropic CEO Dario Amodei urges slower AI development

AI safety researcher Dario Amodei has urged policymakers and industry leaders to adopt stricter oversight mechanisms for advanced artificial‑intelligence systems, warning that the rapid development of large‑scale models could soon enable capabilities that pose a serious threat to global security. In a recent public statement, Amodei highlighted the accelerating pace at which generative AI technologies are being deployed across sectors, arguing that existing regulatory frameworks are insufficient to address the potential for misuse, accidental harm, or coordinated attacks facilitated by these tools.

The call comes as governments and technology firms grapple with mounting evidence that AI models can be weaponized to generate disinformation, automate cyber‑attacks, or create biologically relevant designs with limited oversight. Recent incidents, including the use of text‑to‑image generators to produce realistic fake media and the deployment of language models in phishing campaigns, have intensified debate over the need for mandatory safety testing, transparency requirements, and international coordination. Amodei’s appeal underscores a broader consensus among experts that proactive governance is essential to mitigate the risk of AI‑driven damage on a worldwide scale.