Anthropic says it has disrupted several malicious operations involving its Claude AI models, covering cyber-espionage, weapons development and large-scale surveillance campaigns.
In northern Yemen, the company said it blocked an attempt to use Claude for missile guidance software, including a guided rocket and a long-range ballistic missile. According to its threat report, operators used Claude “in place of human software engineers” and assigned separate model instances to develop missile-guidance and flight-control software.
Anthropic said its safeguards stopped many requests, but some passed through. The operators reportedly concealed their wider objectives and divided tasks across separate sessions to avoid detection. The company said it found no evidence that a working weapon was deployed, although the group appeared to have carried out an unsuccessful test-fire.
Anthropic said it banned the accounts involved and “shared threat information with public- and private-sector partners to mitigate risks posed by the actors”.
The report also linked Claude to a Russian-associated cyber-espionage operation bearing the hallmarks of Midnight Blizzard, or APT29. Anthropic said automated AI workflows handled almost the entire operation, from phishing and setup to data theft targeting Ukrainian, European and diplomatic organisations, including drone makers.
Separately, a Chinese operation involving university students in Hunan province allegedly used Claude “as the engineering and orchestration layer” for attacks on government and corporate networks across the Middle East, Europe and Southeast Asia. Anthropic said it banned the related accounts and added monitoring to detect similar activity.
The company also said it removed 3 Iranian state-aligned accounts involved in covert influence and psychological operations. The operations were linked to Iranian propaganda institutions, including the Islamic Culture and Communications Organisation and a Mashhad seminary command room distributing content aligned with Islamic Revolutionary Guards Corps narratives.
In another case, a China-aligned account “with no Arabic language skills” allegedly used Claude for a “multi-day recruitment operation to infiltrate Uyghur targets in Syria”. Anthropic said, “the model drafted outreach in the regional dialect” and “translated replies in real time”.
Anthropic is also investigating repeated AI safety incidents and has hired an independent research firm to review them.
The company remains in a tense relationship with Washington after the Pentagon designated it a supply chain risk for refusing to remove safeguards against autonomous weapons and domestic surveillance. A California judge later ruled that the designation was unlawful. Despite the dispute, the Pentagon has reportedly used Claude in military missions in Iran and Venezuela.
Also read: Viksit Workforce for a Viksit Bharat
Do Follow: The Mainstream LinkedIn | The Mainstream Facebook | The Mainstream Youtube | The Mainstream Twitter
About us:
The Mainstream is a premier platform delivering the latest updates and informed perspectives across the technology business and cyber landscape. Built on research-driven, thought leadership and original intellectual property, The Mainstream also curates summits & conferences that convene decision makers to explore how technology reshapes industries and leadership. With a growing presence in India and globally across the Middle East, Africa, ASEAN, the USA, the UK and Australia, The Mainstream carries a vision to bring the latest happenings and insights to 8.2 billion people and to place technology at the centre of conversation for leaders navigating the future.