Global / AI Safety
Anthropic says Chinese AI developers secretly used Claude to serve users
State-backed competitors appear to be harvesting Claude's capabilities through illicit access rather than building their own systems.
Anthropic has alleged that China's Moonshot and DeepSeek have secretly used Claude to generate answers served to their own users. The discovery suggests competitors are leveraging Anthropic's technology without permission or transparent attribution.
Published · significance 57 of 100 (medium) · 1 source
What happened
Anthropic has alleged that Chinese AI developers Moonshot and DeepSeek have secretly used Claude to generate responses that were then served to their users. The practice represents illicit access to Anthropic's technology, likely obtained through proxy servers or similar methods that circumvent official channels and terms of service.
Why it matters
The allegation exposes a vulnerability in how frontier models can be harvested and repackaged by competitors, particularly state-backed firms with resources to scale illicit access. It also underscores the tension between open access to AI systems and security: Anthropic's Claude is widely available, but unauthorised commercial use by competitors compounds the challenge of enforcing boundaries around proprietary technology. This pattern—rivals using a leader's model to undercut its market position—could accelerate calls for stronger API controls and monitoring.
What changes
Companies offering frontier models face renewed pressure to detect and prevent unauthorised harvesting at scale. Regulators may scrutinise cross-border access to AI systems, particularly where state-linked firms are involved. Developers and users need clearer visibility into which AI technology actually powers the services they rely on.
Involved
Sources
Written by AI from the reports above; scored by a published formula. How we work. Found a mistake? Email lockedinshreyash@gmail.com. Up To Date summarises and links to original reporting; it never reproduces articles.
Related coverage
- Anthropic details distillation attacks from Alibaba, Moonshot AI and DeepSeek — The escalation underscores the intensifying technical arms race between Western frontier labs and China's AI development. (2026-09-10)
- Anthropic CEO urges AI labs to slow model development for safety oversight — Dario Amodei proposes industry-wide commitment to external safety monitoring, citing risks that outpace control. (2026-09-12)
- Anthropic disrupts Russian and Chinese efforts to misuse Claude for cyberattacks and bioweapons research — The frontier lab is now publishing details of state-sponsored and criminal abuse attempts against its own models. (2026-09-10)
- Anthropic details malicious AI activities across fraud, cyber operations and weapons misuse — The lab disclosed disrupted attacks spanning scams, influence operations and surveillance, raising pressure on frontier labs to demonstrate safety controls. (2026-09-11)
- Anthropic and OpenAI to embed independent safety evaluators in their labs — Companies seek outside oversight to manage risks, but researchers doubt independence without regulatory teeth. (2026-09-16)