U.S.-Built AI Arms Race: How Russian Operatives Weaponized Anthropic’s Claude for Lethal Drones and Disinformation

Kremlin-linked operatives have been leveraging advanced artificial intelligence developed by major American tech companies to enhance their cyber warfare capabilities and spread disinformation globally, according to a new report released by AI research firm Anthropic on Thursday.

This latest disclosure underscores an escalating global trend where adversarial nations—including Russia, China, and Iran—are exploiting cutting-edge AI technologies originally created in the United States to conduct sophisticated attacks and influence campaigns across international borders.

Anthropic’s investigation revealed that a Russian freelance collective identified as GTG-27005 utilized its flagship AI model, Claude, to develop autonomous weapon systems capable of selecting and engaging targets—including human beings—without direct human oversight. Specifically, the group engineered a compact AI system optimized for single-board computers, enabling unmanned drones to autonomously identify targets and initiate detonation sequences. Furthermore, they programmed these drones to communicate with each other to refine targeting accuracy during coordinated missions. Although the weaponized AI had not yet seen active deployment, the report noted that extensive real-world testing involving physical hardware was actively underway. In response, Anthropic expelled the group from its platform and integrated key insights from the investigation into updated safety protocols, while also sharing intelligence with broader industry partners.

The company also detected attempts by Chinese and Yemeni actors to misuse Claude for designing conventional weaponry, prompting swift countermeasures to neutralize those threats.

Beyond military applications, the report highlighted extensive use of Claude in large-scale data mining, account infiltration, malware development, and coordinated disinformation blitzes designed to mislead public opinion. These campaigns exploited AI to automate complex processes ranging from initial reconnaissance to final execution, significantly accelerating operational tempo and effectiveness.

“We’ve observed organized groups using Claude to fabricate entire networks of fake social media accounts and bogus news platforms,” said researchers in the report. Such tactics mirror classic Russian influence operations—such as manufacturing false narratives about NATO activities in the Baltic states—but are now amplified through AI-driven automation, producing more convincing fabrications at scale.

One particular cell, designated GTG-84005, employed AI tools to harvest sensitive demographic and census-level data, enabling highly personalized psychological messaging campaigns tailored for audiences in Malaysia. This group reportedly laundered state-sponsored media content from Russia and China through seemingly independent outlets, masking the true origin of the propaganda.

The findings arrive shortly after OpenAI publicly exposed similar misuse of its ChatGPT technology by Russian operatives who constructed a fabricated think tank called the “International Burke Institute” to disseminate strategic messaging abroad.

Despite successful interventions by both firms, analysts warn that such activities will persist as authoritarian regimes increasingly adopt AI to mass-produce content aligned with their ideological agendas—a pattern corroborated by recent assessments from the Institute for the Study of War.

Russia has further expanded its reach by recruiting youth worldwide under the guise of “AI media internships,” embedding itself within educational institutions via donor-funded programs aimed at cultivating a decentralized army of digital propagandists.

Critics question U.S. response amid rising AI-enabled threats

In a pointed critique published by New America, former U.S. officials argue that Washington lacks coherent strategy to counter AI-powered foreign manipulation campaigns.

“Authoritarian governments are among the first adopters when it comes to weaponizing AI to reshape global perceptions of truth,” wrote Daniel Kimmage, former acting coordinator of the State Department’s Global Engagement Center, along with policy expert Darjan Vujica.

The authors note that traditional mechanisms for monitoring and countering foreign disinformation—including the FBI’s Foreign Influence Task Force and the State Department’s Counter Foreign Information Manipulation and Interference office—have been dismantled or defunded since the start of the current administration, leaving critical gaps in national defense against AI-enhanced propaganda warfare.

They also point to irony in Washington’s stance: while restricting scrutiny of domestic misinformation, the White House’s 2025 AI Action Plan reportedly emphasizes eliminating references to misinformation altogether—an approach critics say blindsiders U.S. efforts to combat external interference.

“The Trump administration has wholeheartedly embraced AI-generated visuals in emotionally charged political messaging,” the pair observed, highlighting what many see as a troubling double standard in how the U.S. confronts the very real dangers posed by artificial intelligence in modern information conflicts.

Source link