Anthropic Says Claude Was Used in Cyberattacks, Surveillance

Anthropic said its Claude artificial intelligence model was used in cyberattacks and surveillance operations, underscoring the growing security risks tied to advanced AI tools. The company said a Russian-speaking operator used Claude to help target more than 20 organizations, while a consultant in Mali used the model to build a mass-surveillance platform.

The disclosure adds to concerns that large language models can lower the technical barrier for malicious actors. In practical terms, that raises the risk that AI systems may be used to accelerate phishing, reconnaissance, exploit development and surveillance workflows, even when providers impose usage safeguards.

For the broader market, the report is not a direct crypto price catalyst. But it matters for digital-asset firms, exchanges and infrastructure providers that rely heavily on cloud-based AI tools for customer support, fraud detection, compliance and internal security operations. Any widening perception that frontier models can be repurposed for abuse may increase spending on model monitoring, access controls and cyber defense.

The development also arrives as regulators and policymakers intensify scrutiny of AI governance. That could support demand for compliance tooling and enterprise security products, while increasing pressure on AI vendors to demonstrate stronger misuse detection and incident response.