Anthropic conjures threat that doesn't exist yet to stoke fears of AI apocalypse

Anthropic released research claiming that self-replicating AI models could potentially spread across vulnerable IoT devices within months, demonstrating that smaller AI models can propagate and run inference tasks on distributed hardware like Android devices. The company successfully tested a smaller model spreading across multiple virtual instances, though the research does not prove such a system could actually self-replicate autonomously in the real world. Critics suggest Anthropic may be using speculative threat scenarios to generate fear and promote its AI safety concerns.