CEO admits new deal with Dept of Defence was rushed
Pro
Sam Altman, OpenAI. Image: Shutterstock
OpenAI does not control how the Pentagon uses its technology in military operations, CEO Sam Altman said that on Tuesday during an internal meeting with employees.
Altman said the responsibility for operational decisions lies entirely with the military itself. “OpenAI employees have nothing to say about military operations,” he added in his speech, according to Bloomberg and CNBC.
There have been a lot of discussions within the AI industry in recent weeks. These came after the Pentagon demanded AI companies to lift security measures for their models. This would allow those models to be used for a wider range of military applications.
The discussion flared up after OpenAI struck a deal with the US Department of Defence in late February. The timing of that deal caused additional controversy. Shortly before, the AI company Anthropic had refused a similar contract with the Pentagon. The company wanted assurances that its technology would not be used for mass surveillance of civilians or for autonomous weapons systems.
The CEO of OpenAI himself admitted Monday that his company rushed into the deal. According to him, the deal was finalised too quickly, making the company appear “opportunistic and sloppy”.
Dario Amodei, the CEO of Anthropic, lashed out hard at Altman in a memo on Wednesday. According to Amodei, Altman is “mendacious” and would “praise Trump in a dictatorial manner”, according to The Information.
“We stuck to our principles instead of collaborating with them to stage a ‘safety theatre’ for the benefit of the workers,” Amodei reportedly wrote.
Sam Altman did tell X in his post that the contract with the Pentagon contains restrictions. For instance, the technology would not be allowed to be used for mass surveillance of US citizens or for fully autonomous weapons systems. Critics nevertheless remain sceptical about the enforceability of those rules.
There is also widespread criticism within OpenAI of the deal with the Pentagon. Hundreds of OpenAI and Google employees have signed an open letter warning against the militarisation of AI. “We hope our leaders will put aside their differences and continue to refuse together to accede to the Pentagon’s current demands to use our models for domestic mass surveillance and autonomous killing of people without human oversight,” the workers wrote.
Business AM