Snapshot from Nov 22, 2026 at 07:00 UTC. For live data and tracking: View Live
Tech cybersecurity report

Anthropic Threat Report Flags Claude Misuse

Analysis based on 86 articles · First reported Sep 10, 2026 · Last updated Sep 14, 2026

Sentiment
-25
Attention
6
Articles
86
Market Impact
General
Live prominence charts, article sentiment distribution, and event development timeline available on the Ergen Dashboard

The report reinforces regulatory and reputational risk narratives around frontier AI models, potentially strengthening the case for tighter oversight and compliance costs across the sector. It also highlights rising demand for AI safety, threat intelligence and cybersecurity spending, while raising competitive and legal risk for the named Chinese AI labs and for Anthropic itself.

Artificial Intelligence Cybersecurity Defense

Anthropic published its latest Threat Intelligence report, a roughly 150-page document covering malicious use of its Claude models over about eight months from December 2025 to August 2026. The company said it disrupted five cases in which scientists used Claude in ways that could support biological weapons development, including planning of avian influenza mammalian-adaptation experiments, chikungunya and orthopoxvirus-related work, and it banned the accounts involved. It also documented six cases of conventional weapons development, including a weapons cell in northern Yemen that used Claude Code to write guidance, navigation and control software for a guided rocket, a multi-stage ballistic missile with a claimed range over 2,000 km, and hypersonic glide vehicle variants; the cell reportedly test-fired a guided rocket that failed. Anthropic attributed a suspected Russia-linked cyber espionage campaign against Ukrainian government, military and diplomatic targets to tradecraft consistent with Cozy Bear, and said it disrupted activity linked to ShinyHunters affiliates. It accused seven China-based labs, including Alibaba, Moonshot, DeepSeek and Xiaomi, of distillation and capability-extraction attempts, with Alibaba allegedly running the largest illicit distillation operation involving more than 151 million exchanges between May and July 2026. The report also covered state-linked surveillance and influence operations in Iran, China, Mali, Russia, France, Malaysia and the United Arab Emirates. China's foreign ministry said it was unaware of the report and opposed what it called distortion of facts and smears.

90 Anthropic identified and disrupted account Iran
90 Anthropic published threat intelligence report
85 Anthropic shut down surveillance operations
80 Anthropic disrupted cyber espionage campaign Ukraine
80 Anthropic disrupted illicit distillation attacks Alibaba Group
80 Anthropic published threat report
80 Anthropic blocked biological research requests
78 Houthis seized control Yemen
77 Anthropic disrupted attacks Alibaba Group
75 Houthis developed missile guidance software
75 Saudi Arabia carried out airstrikes Houthis
75 DeepSeek routed customer conversations Anthropic
75 Cozy Bear ran cyber espionage Ukraine
+ 80 more actions View on Dashboard
priv
Anthropic is the author of the report and the central actor, disclosing misuse of its Claude models and detailing its safeguards and enforcement actions. The findings raise regulatory scrutiny and reputational risk but also position it as a leader in AI safety and threat intelligence.
Importance 100.0 Sentiment -10.0
stock
Anthropic accused operators linked to Alibaba of running the largest illicit distillation attack, with over 151 million exchanges between May and July 2026 aimed at improving its Qwen models. The allegation raises legal, reputational and potential regulatory risk for the company.
Importance 70.0 Sentiment -35.0
mil
A Yemen-based weapons cell matching the Houthis reportedly used Claude Code to develop missile guidance software and test-fired a guided rocket that failed. The disclosure highlights the group's weapons ambitions and the proliferation risk of AI tools.
Importance 65.0 Sentiment -20.0
cnt
China-based labs and state-linked actors were accused of distillation attacks, surveillance and weapons-related work; Beijing rejected the report as a smear.
Importance 60.0 Sentiment -20.0
cnt
Iran-linked operators allegedly used Claude to compile targeting handbooks on US naval assets and ran influence operations tied to state institutions. The activity adds to geopolitical tensions with the United States.
Importance 55.0 Sentiment -25.0
cnt
Russia-linked actors were cited for weapons development, propaganda and the Cozy Bear espionage campaign against Ukraine, increasing sanctions and cyber risk perceptions.
Importance 55.0 Sentiment -25.0
priv
Anthropic alleged Moonshot routed live customer conversations through Claude and used the responses as training data, raising privacy and compliance concerns for the Kimi chatbot creator.
Importance 55.0 Sentiment -30.0
priv
Anthropic alleged DeepSeek similarly routed live customer conversations through Claude for training data, adding to scrutiny of Chinese AI developers' practices.
Importance 55.0 Sentiment -30.0
cnt
Northern Yemen hosted the weapons cell that used Claude for missile software, drawing attention to the conflict zone's weapons programs.
Importance 45.0 Sentiment -15.0
cnt
US naval assets and bases were reportedly targeted by Iran-linked intelligence gathering, and US policymakers are debating AI regulation.
Importance 40.0 Sentiment -10.0
cnt
Ukrainian government, military and diplomatic officials were targets of the alleged Russia-linked AI-assisted espionage campaign.
Importance 40.0 Sentiment -10.0
cnt
Saudi Arabia is cited as facing continued Houthi attacks on its oil installations and conducting retaliatory air strikes, keeping regional energy risk elevated.
Importance 35.0 Sentiment -20.0
stock
Xiaomi was named among seven China-based labs whose attacks Anthropic said it disrupted, exposing it to reputational and regulatory risk.
Importance 35.0 Sentiment -20.0
oth
Russia-linked threat actor conducted cyber espionage against Ukraine using AI.
Importance 30.0 Sentiment -20.0
ngo
Importance 30.0 Sentiment 0.0
+ 45 more entities View on Dashboard
Anthropic related ↔ Alibaba Group
Anthropic related ↔ Houthis
Anthropic related ↔ China
Anthropic related ↔ Iran
Anthropic related ↔ Russia
Anthropic related ↔ Moonshot AI
Anthropic related ↔ DeepSeek
Anthropic related ↔ Yemen
Anthropic related ↔ United States
Alibaba Group related ↔ China
Alibaba Group related ↔ Iran
Alibaba Group related ↔ Russia
Alibaba Group competitor ↔ DeepSeek Alibaba Group is a major Chinese tech company that directly competes with DeepSeek, a Chinese AI startup, in the develop
Alibaba Group related ↔ Yemen
+ 27 more relationships View on Dashboard
ERGEN INTELLIGENCE
Track this event live

Set up alerts, explore entity relationships, search across thousands of events, and build custom intelligence feeds.

Open Dashboard

About Ergen

Ergen is a news intelligence platform that converts raw news articles into structured data. It tracks events, entities, and the relationships between them, with sentiment and attention metrics derived from thousands of articles. Pages on this site are daily static snapshots from the platform's live database. For real-time tracking, search, and alerts, the full dashboard is at app.ergen.ai.