Anthropic warns of near-term AI replication risk on IoT devices Anthropic's experiments reveal that while small AI models are not yet fully adept at autonomous replication, they are advancing rapidly. The firm predicts that these models could become capable of self-reproducing across commercial IoT devices by the end of the year, posing a significant security risk due to the lack of a central shutdown mechanism and the models' ability to adapt to unfamiliar hardware. In simplified tests, only GLM-5.2 completed full replication across a test network, while other tested open-weight models failed, Anthropic says. Anthropic warns replicating AI would have no central switch-off point and could figure out unfamiliar devices on its own. (Reuters) Anthropic said on Monday its experiments suggested small AI models may be months away from being capable enough to self-replicate on commercial internet-of-things (IoT) devices, such as smart ovens and off-brand Android TV boxes. In its experiments, Anthropic looked at whether AI models could find vulnerable devices, copy themselves into those devices, and continue spreading without human help.GLM-5.2 Completes AI Model ReplicationHowever, the firm found that such AI models also struggled to complete the entire sequence from start to finish. In its simplified tests, where some security steps were removed, Anthropic found that GLM-5.2 and stronger models were capable of end-to-end replication.IMF warns AI market correction, cyberattacks threaten global financial stabilityReplicating models would be hard to shut down as there is no central switch-off point and, unlike non-AI malware, they could also figure out unfamiliar devices on their own, targeting data centre processors, Anthropic warned. The firm focused on IoT devices like Android TVs, smart appliances, and other internet-connected gadgets in its research as these are typically poorly secured.AI Replication Timeline And Security WarningsThe AI major also warned that if small models improved at their historical rate, they would be able to autonomously replicate by the end of this year, with small weight-based replicators following suit soon after. Anthropic also said such autonomous AI systems were likely to replicate slowly and rely on widely available, relatively insecure computing resources. It added that systems should be strengthened to make it harder for such AI agents to spread.IMF warns AI market correction, cyberattacks threaten global financial stabilityIMF chief warns AI could disrupt jobs, widen Asia's economic divideRSF drone attack destroys hospital in Sudan's Blue Nile, injures two medical workersUN global poverty index rises for first time, affecting 1.2 billion peopleSaudi Arabia commits to uninterrupted energy supplies at ministerial forum in RiyadhSaudi, Iranian foreign ministers discuss latest war escalation