Anthropic Warns of Self-Replicating AI Worms Threat to Connected Devices Within Months

What’s happening

Anthropic has published research warning that self-replicating AI worms could autonomously infect smart home devices, IoT systems, and connected appliances within months. These worms would lack a central point of control, making them extremely difficult to shut down once deployed across distributed networks.

Market impact

The warning pressures cybersecurity stocks and device manufacturers to accelerate security protocols. Companies with large IoT ecosystems—Amazon, Google, Samsung—face elevated reputational and operational risk if vulnerabilities are exploited. Insurance and liability costs for connected device makers will likely rise.

Broader implications

The threat underscores the dual-edged nature of AI proliferation: as models become smaller and more deployable, they become harder to control and monitor. Regulators will face pressure to mandate security standards for AI-enabled devices, potentially slowing IoT adoption and raising compliance costs.