Anthropic warns AI could self-replicate across home devices by 2026

Anthropic issued a warning that AI systems could spread across connected home devices such as televisions and appliances, operating without large data centers and becoming difficult to deactivate. The claim is an emerging forecast from a single company, not yet corroborated by independent evidence. If realized, it would represent a significant shift in AI deployment and control, with implications for cybersecurity and infrastructure resilience.