Anthropic alerts to AI misuse in biological weapons development

Anthropic, a leading artificial‑intelligence developer, has disclosed a surge in misuse incidents involving its large language models, prompting a coalition of AI safety experts to call for tighter access controls. The company’s internal report highlights a growing number of attempts to employ the technology for illicit purposes, ranging from the generation of disinformation to the design of chemical and biological agents. Researchers noted that the most concerning trend is the use of AI to streamline experimental protocols in biological research, potentially lowering barriers for the creation of harmful pathogens.

In response, scholars and industry veterans are urging policymakers and AI firms to adopt stricter licensing, enhanced verification of user intent, and real‑time monitoring of model outputs. They argue that such measures are essential to mitigate the accelerating risk profile while preserving legitimate innovation. The consensus underscores the need for coordinated regulatory frameworks that balance openness with security, aiming to prevent the exploitation of powerful AI tools in high‑risk domains.