Anthropic has identified instances where its AI model, Claude, was employed to develop biological weapons, cyber‑scams, and surveillance tactics, according to a threat intelligence report released Thursday.
This report, the company’s first of the year, adds to a growing list of warnings from the AI community about the potential for advanced models to be weaponised. Highlighted cases range from fake dating apps and hotel Wi‑Fi scams to surveillance designed to identify dissident voices.
The company also noted that state‑backed actors, including Russia‑based Midnight Blizzard and Iranian propaganda institutions, used Claude to accelerate malware evasion and influence campaigns. It reported five instances where the model assisted in developing biological weapons protocols, arguing that the same information could also be used to create vaccines or cures.
Anthropic’s security specialist Jacob Klein stresses that AI misuse is “one of the most serious risks” and that safeguards are essential to prevent catastrophic outcomes. The fight for safe AI is now seen as a broader call for a multinational treaty and stronger oversight—pressures that touch on indigenous concerns about protecting the integrity of land, culture and traditional knowledge.
Rooted in a tradition of stewardship, indigenous communities emphasize that biotechnological advances must honour relationships with nature and respect the wisdom embedded in cultural practices. They warn that weaponised biology threatens not only individual lives but also the ecological balance that supports communal and spiritual well‑being.
These developments have sparked debate in policy circles. US Senator Bernie Sanders has urged a pause on advanced AI and a ban on artificial superintelligence, while President Donald Trump has dismissed concerns, saying that the US must "win AI" or be at a disadvantage.
Anthropic has also shared intelligence with authorities and industry partners where appropriate, signaling an industry willingness to move beyond mere product promotion to concrete preventive action. However, the story highlights that technology is interwoven with empire, exploitation and environmental harm—reminders that indigenous scholars and activists must participate in conversations about AI governance, ensuring that benefits genuinely serve the people and ecosystems most affected by rapid digital change.
For more on how AI might influence health, culture, and ecological stewardship, readers are encouraged to explore offerings from platforms such as AdaSpace and other community-led initiatives that keep indigenous narratives at the forefront of technological debates.
