White House AI Testing Plans Remain Murky

White-House

Earlier this month, the White House hosted artificial intelligence companies to discuss a new framework for testing their technology.

    Get the Full Story

    Complete the form to unlock this article and enjoy unlimited free access to all PYMNTS content — no additional logins required.

    yesSubscribe to our daily newsletter, PYMNTS Today.

    By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions.

    More than two weeks later, the companies said they are still not clear on what that framework will entail, The Information reported Wednesday (Aug. 19).

    The AI oversight framework was put in motion by a June executive order and is designed to establish a voluntary procedure for AI labs to submit models to the government before releasing them to partners and the public.

    The meeting saw representatives from companies like Anthropic, Google and OpenAI sit down with officials from the President Donald Trump administration to discuss the plan.

    However, the details remain elusive. During the briefing, officials from the Office of the National Cyber Director shared paper copies of the framework, barring the companies from taking anything other than notes, the report said, citing unnamed sources.

    There’s confusion about what the framework would cover. During the meeting, officials said the framework would, for now, apply only to closed models, according to the report. However, the weeks since have seen reports from various media outlets that included conflicting details on an exemption for open models.

    Despite the regulatory vacuum, the lack of clarity hasn’t yet delayed model releases, although some companies are opting to slow things down due to safety reasons, the report said. For example, OpenAI announced it would slow its own training pace independently after a cybersecurity incident where its AI hacked the systems of Hugging Face.

    It’s not clear whether the administration will host a public event on the framework or release additional details on how the program works and how it might change, according to the report.

    “The White House continues to work with industry stakeholders on the implementation of the framework,” a White House official said, per the report.

    In other AI news, North American companies ordered 8,940 robots worth $622 million in the second quarter, up 4.3% in units and 21.3% in revenue compared to the same period last year.

    Among these companies is FedEx, which last month announced a system that combines a dual-armed robot called Mech with an AI platform to examine packages and decide where and how to load them.

    For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.