Anthropic posted a job advertisement seeking a policy manager to deal with conventional weapons, just two days after its CEO warned that creating an advanced intelligent system could lead to human extinction.
On September 12, 2026, Dario Amode, CEO of Anthropic, published an article urging advanced AI companies, including his own, to slow down development for fear of destroying human life in ways we cannot even imagine.
Just two days later, Anthropic posted a job advertisement on Daybook, a platform specializing in jobs related to policy and government affairs, looking for a policy design manager for conventional weapons, to work within the "safeguards" team tasked with defining "limits imposed on how Claude is used."
The job posting explains that the person who will fill this position will be responsible for determining when the use of artificial intelligence in conventional weapons is acceptable and when it crosses the line.
The announcement stated: "It is difficult to draw a clear line between acceptable and harmful requests here, because the basic components can be used for more than one purpose: the same capabilities that serve civil engineering and scientific research can also be used in making weapons systems."
The announcement adds: "Conventional weapons are increasingly reliant on software, and the risks are not limited to firearms: from an AI model operating a weapon system or writing guidance code, to arming platforms that can be used both civilian and military. This role extends to all types of weapons, even autonomous systems that select and target objectives without human intervention."
Anthropic maintained its hardline stance on autonomous weapons, severing ties with the Pentagon when it refused to allow the Department of Defense to use its prototypes in weapons that operate without human intervention. However, its position was far more flexible regarding weapons that still require someone to press a firing button.
According to Reuters, Anthropic tools were used in the US attack on Iran, and the technology was also allegedly used in the US raid on Venezuela that led to the arrest of President Nicolas Maduro, according to the Wall Street Journal.
Oddly, Amoudi’s alarming letter didn’t address weapons systems, whether autonomous or otherwise—uses where AI is already being integrated to kill people. Instead, it spoke of hypothetical risks that haven’t yet materialized, such as internet-disrupting hacks or the development of biological weapons. Anyone who fears AI could kill everyone would logically expect their company to take a firm and unequivocal stance against conventional weapons, not just adopt a policy of appeasement that allows it to reconcile a claim of responsible development with maintaining lucrative defense contracts.
what is your opinion on this news ?do let us know in the comments on tha comment box
if you liked this news article please share on friends & family & don't forget to follow on website & social media
