Just one day after it was revealed that the Donald Trump administration was considering tightening oversight of the artificial intelligence sector, Google, Microsoft, and xAI have agreed to provide the US federal government with early access to the AI systems they develop. According to The Wall Street Journal, the Artificial Intelligence Standards and Innovation Center (CAISI), part of the Department of Commerce, will evaluate these companies' new models.
Chris Fall, Director of CAISI, stated that independent and rigorous measurement methods are critical to understanding the national security implications of advanced artificial intelligence. According to Fall, such collaborations help expand the scope of work carried out for the public good.
Government to Access Models with Fewer Restrictions and Without Security Measures
Looking at the details of the agreement, it is stated that companies will provide their developed models to CAISI with fewer restrictions, and even with some security measures disabled. This will allow the institution to more thoroughly test the national security risks and capabilities of these systems.
This development comes immediately after news that the Trump administration plans to introduce new regulations for artificial intelligence. The White House is reportedly considering establishing a working group that will oversee future AI models and have the authority to review them before they are released to the public.
At first glance, this step might seem like a retreat from Trump's previously announced free-market approach to AI policy. However, some commentators suggest that the administration has been trying to steer the sector according to its own priorities from the outset.
Indeed, the tension between the Pentagon and Anthropic is seen as part of this process. The Department of Defense had attempted to label Anthropic as a supply chain risk due to the company's insistence on security measures preventing its technology from being used for mass surveillance and autonomous weapons.
"Woke" AI Systems Banned
Furthermore, Trump's executive order, titled "Preventing 'Woke' Artificial Intelligence in the Federal Government," signed on the day he announced his AI action plan, also draws attention. This decision prohibits federal agencies from using AI systems that "guide responses in a way that promotes ideological approaches such as DEI." This move effectively creates an ideological compatibility test for companies, and Anthropic is believed to have failed this test. For this reason, many companies are thought to have chosen to cooperate with the government to avoid a similar situation.
0 Comments: