> ## Content Index > Fetch the complete content index at: https://www.implicator.ai/llms.txt > Use this file to discover other available public pages before exploring further. # Anthropic Details Claude’s Role in Surveillance in Mali, China and Iran - URL: https://www.implicator.ai/anthropic-claude-surveillance-mali-china-iran/ - Published: 2026-09-11T11:24:24.000Z - Updated: 2026-09-11T11:24:24.000Z - Description: Anthropic says Claude helped build surveillance systems and produce dossiers in Mali, China and Iran. Its report shows why account bans can have different effects once software is deployed. The systems’ reported reach and downstream harm could not be independently verified. - Author: Marcus Schuler - Tags: Politics Anthropic said Thursday that it had blocked Claude accounts tied to state-aligned actors and contractors who used its models to help build or run [surveillance operations in Mali, China and Iran](https://www.anthropic.com/threat-intelligence-report-september-2026?ref=implicator.ai) between January and July 2026\. The activity ranged from engineering a locally deployed national interception platform to producing dossiers on religious leaders, dissidents and diaspora communities. All the cases involved Haiku, Sonnet or Opus, rather than the newer Fable or Mythos models. AI is reducing the staff and specialist knowledge needed for intelligence work, said Jacob Klein, Anthropic’s head of threat intelligence. “[They’re effectively automating parts of the job within the intel apparatus](https://www.axios.com/2026/09/10/anthropic-claude-government-surveillance-threats?ref=implicator.ai),” Klein said. The findings are based on Anthropic’s review of activity on its own services and its attribution of the operators. This article could not independently verify the reported systems’ reach or downstream harm. The company’s broader report covers cases it investigated from December 2025 through August 2026. What the Report Found - Anthropic linked Claude accounts to surveillance operations in Mali, China and Iran between January and July 2026. - Anthropic put Mali’s claimed surveillance scope at roughly 25 million SIM cards. Its account ban did not disable the locally deployed system. - Two Iranian units used Claude to develop tools and analyze data feeding a shared case-management system called Arman. - A commercial surveillance project was detected during its pilot. The reported systems’ reach and downstream harm could not be independently verified. AI-generated summary, reviewed by an editor. [More on our AI guidelines](https://www.implicator.ai/about/). ## Mali system survived the account ban One subscriber, assessed as likely to be an independent consultant in Bamako working for Mali’s Agence Nationale de la Sécurité d’État, used Claude as the main engineering resource for a system called Lakana 360\. Anthropic said the platform was built to monitor roughly 25 million SIM cards across Mali’s three national mobile operators. That figure describes SIMs within the system’s claimed scope, not people or confirmed victims. Claude helped design software that could collect call records, messages and voice traffic, link identities across SIM cards and generate a dossier for a phone number. Anthropic said the consultant removed a warrant check from the dossier function at an operator’s request. The platform ran locally on other AI models. Banning the Claude account interrupted further design work but did not disable the deployed product. FREE WEEKDAY MORNING BRIEFING Follow how AI changes surveillance. The Implicator Morning Briefing filters the AI news cycle to the stories worth your attention and explains their consequences. From San Francisco, every weekday at 4:45 a.m. Pacific, 7:45 a.m. Eastern. Email address Send me tomorrow’s briefing Check your inbox for the confirmation link. About five minutes. No hype. No spam. [Human Rights Watch documented](https://www.hrw.org/world-report/2026/country-chapters/mali?ref=implicator.ai) arbitrary arrests, disappearances and a widening crackdown on opposition and media in Mali during 2025. ## China’s casework becomes a pipeline The China-linked activity covered religious groups, domestic critics and communities abroad. One operator used Claude to turn multilingual material into Chinese-language dossiers on Catholic leaders, Tibetan Buddhist civil society, Falun Gong practitioners and Taiwanese Christians. The files included personal histories, social media accounts and potential pressure points. Other accounts produced government-style briefings that rated reporting and online posts for political sensitivity. One cluster was tied to municipal public and state security work with varying levels of confidence. A separate operation was assessed with medium confidence to be a contractor serving government clients, while an operation targeting Uyghurs in Syria was assessed with low confidence to involve a contractor rather than a state office. In that Syria case, a Chinese-speaking operator with no Arabic skills used Claude for translation and message review during an attempted recruitment effort. The model also structured material collected elsewhere into target profiles. It refused some requests, including covert interrogation, but other surveillance tasks passed its safeguards. ## Iranian units feed Arman Anthropic banned 16 accounts operated by two linked Iranian units that used Claude for software development and analysis. The company assessed with high confidence that the units were associated with paramilitary and domestic security entities. Know someone who'd find this useful? [✉️ Email it to a friend in one click](mailto:?subject=A%20newsletter%20I%20think%20you%27d%20like&body=This%20is%20one%20of%20maybe%20three%20newsletters%20I%20actually%20read.%20The%20rest%20just%20pile%20up%2C%20unread%2C%20judging%20me.%0A%0AAnd%20yes%2C%20this%20email%20mostly%20wrote%20itself%2C%20which%20is%20a%20little%20on%20the%20nose%20for%20an%20AI%20newsletter.%20Doesn%27t%20make%20it%20wrong.%20implicator.ai%20is%20good.%0A%0ASubscribe%20free%3A%20https%3A%2F%2Fwww.implicator.ai%2Fsubscribe%2F%3Futm%5Fsource%3Dnewsletter%26utm%5Fmedium%3Dforward%26utm%5Fcampaign%3Demail%5Fforward), or they can [subscribe free here](https://www.implicator.ai/subscribe/?utm%5Fsource=newsletter&utm%5Fmedium=forward&utm%5Fcampaign=forward%5Fto%5Fcolleague). Both units fed a case-management system called Arman. One built a web interface for the system and analyzed social media activity, selecting 39 opposition and diaspora accounts for monitoring. The other used Claude to develop a browser extension that collected identities from social networks. Anthropic said safeguards refused explicit profiling and propaganda requests but did not refuse many surveillance-software tooling requests. Arman’s records included national IDs, beliefs, criminal histories and an “action” tab. ## What the bans stopped The report also described a commercial platform linked to S2T Unlocking Cyberspace that classified users in Iran and the Persian Gulf by location, demographics and political views. Anthropic found the project during a pilot and said it saw no evidence that later stages were used against real targets before the account was banned. That absence is not proof that no downstream targeting occurred. A [2023 investigation of an S2T-linked brochure](https://forbiddenstories.org/osint-s2t-unlocking-cyberspace-journalists-activists/?ref=implicator.ai) described fake accounts, phishing and device compromise as parts of a proposed surveillance chain, but it did not validate the operations disclosed this week. “Authoritarian states are using AI for surveillance, repression and influence operations today,” Klein said. “It’s no longer hypothetical.” Frequently Asked Questions What did Anthropic disclose? Anthropic said accounts linked to state-aligned actors and contractors used Claude to help build or run surveillance operations in Mali, China and Iran between January and July 2026\. The cases involved Haiku, Sonnet and Opus models. Did banning Claude stop Mali’s surveillance platform? Anthropic said its ban interrupted the developer’s design work but did not disable Lakana 360\. Claude had helped engineer the software; other AI models ran the deployed system locally. Does the Mali figure mean 25 million people were monitored? No. Roughly 25 million refers to SIM cards within the system’s claimed scope across three national mobile operators. It is not a count of people or confirmed victims, and the reported reach could not be independently verified. How was Claude used in China and Iran? China-linked operators used Claude to produce dossiers and government-style briefings. Two Iranian units used it for software development and analysis feeding Arman, a shared case-management system. Anthropic banned 16 accounts associated with those units. Did Anthropic identify any surveillance activity before deployment? Anthropic said it found an S2T-linked commercial project during a pilot and saw no evidence that later stages were used against real targets before the ban. That absence does not prove no downstream targeting occurred. AI-generated summary, reviewed by an editor. [More on our AI guidelines](https://www.implicator.ai/about/). ## Related stories [Anthropic AI Policy: Serves Spies, Blocks CopsAnthropic offers classified AI models to intelligence agencies while blocking FBI and law enforcement contractors. Inside the selective policy strategy.Implicator.ai![](https://www.implicator.ai/content/images/2025/09/2025-09-17-04.55.57-anthropic_government@2x.webp)](https://www.implicator.ai/anthropic-delivers-classified-ai-to-spies-bars-fbi-surveillance/) [Iran AI Sanctions Workaround; Claude Retakes LLM Meter; VastIran uses Western AI for cyber ops while building a domestic platform. Claude retakes the LLM Meter lead. Vast raises nearly $200M for Tripo 3D models.Implicator.ai![](https://www.implicator.ai/content/images/2026/06/2026-05-31-21.57.38-nl_2026_06_01_header@2x.webp)](https://www.implicator.ai/iran-finds-the-ai-workaround-washington-cannot-sanction/) [Anthropic AI: Yes to Spies, No to Domestic SurveillanceAnthropic builds classified AI models for intelligence agencies while blocking FBI and ICE from domestic surveillance use. The contradiction has sparked White House tensions and raises questions about ethical lines in government AI.Implicator.ai![](https://www.implicator.ai/content/images/2025/09/2025-09-17-04.56.11-anthropic_government_2@2x.webp)](https://www.implicator.ai/anthropic-draws-a-line-between-spies-and-cops/)