Sam Altman, CEO of OpenAI, admitted during a meeting with company employees that he cannot control how the U.S. Department of Defense (Pentagon) utilizes the company's artificial intelligence technologies, according to a report by CNBC.
اضافة اعلان
Altman clarified that his company does not have the right to make operational decisions based on direct Pentagon requests. He added: "You might think the strike on Iran was good and the invasion of Venezuela was bad. You don't have the right to voice an opinion on that."
Despite this, the report indicates that the Pentagon will allow OpenAI to build the security safeguards it deems appropriate. However, this does not extend to final operational decisions, which rest solely with the Secretary of Defense, Pete Hegseth.
Internal Criticism and Industry Rivalry
Altman faced widespread and public criticism from a group of his employees, both inside and outside the meeting. This backlash follows OpenAI replacing Anthropic after U.S. President Donald Trump ordered a halt in dealings with the latter.
In a memo to his employees, Dario Amodei, CEO of Anthropic, criticized OpenAI’s direct cooperation with the Pentagon, labeling the company's statements as "lies," according to a report by TechCrunch.
Amodei's criticism of Altman intensified in a separate report by The Guardian, where he called Altman a "liar" and accused him of praising Trump in a "dictatorial fashion." At the end of his memo, Amodei asserted that OpenAI’s announced security measures are merely a "theatrical performance" to appease employees—a path he claimed he refused to take.
Background: AI in Military Operations
It is worth noting that Anthropic’s tools were among the first AI technologies integrated into the Pentagon's operations, granting access to classified files. "Claude," the tool developed by the company, reportedly played a pivotal role in planning attacks on Iran and Venezuela. (CNBC)
