Facebook owner Meta Platforms will buy artificial intelligence chips from Advanced Micro Devices in a deal that will also give it the opportunity to buy up to a 10% stake of the chip company.

Meta will buy AMD’s latest chips, the MI450, to help power data centers. The 6-gigawatt agreement will see shipments supporting the first gigawatt deployment set to start during the second half of this year. The agreement could potentially be worth more than $100 billion. 

Shares of AMD finished Tuesday's regular trading session up nearly 9%.

News of the AMD deal comes just days after Meta announced a long-term partnership where it will use millions of chips and other equipment from Nvidia for its artificial-intelligence data centers.

AMD is looking to keep pace with Nvidia, which carved out an early lead in tailoring its GPU chips to train powerful AI systems. As part of Monday's deal, AMD issued Meta a performance-based warrant for up to 160 million shares of its common stock at $0.01 a piece, structured to vest as long as certain milestones are achieved.—The Associated Press

Anthropic has until Friday to agree to the Pentagon's terms about how the military can use its AI tools—or face a variety of potential punishments, including being placed on a blacklist. 

The ultimatum comes after Anthropic CEO Dario Amodei met U.S. Secretary of Defense (err, War!) Pete Hegseth on Tuesday to try to resolve the disagreement, according to Axios, which cited anonymous sources.

A quick recap on the standoff: Amodei says Anthropic's Claude product (the only commercial AI product that has had access to DoD classified systems) cannot be used to attack enemy combatants or to spy on Americans. Hegseth insists the military be able to use Claude, and any other AI company's products, for "all lawful purposes." 

The meeting was cordial but "not warm and fuzzy at all" according to Axios. Neither side budged on the key issue, and Hegseth laid out his ultimatum. If Anthropic does not acquiesce by Friday, the Pentagon could invoke the Defense Production Act and force the recalcitrant AI company to comply. The Pentagon also threatened to declare Anthropic a supply chain risk—a designation that would not only end Anthropic's business with the military, but require any other companies with military contracts to stop using Anthropic.

For Anthropic, valued at $380 billion and considered a top IPO candidate, the threats, if carried out, could have big consequences on its business. Anthropic's commitment to AI safety is a big part of its identity, but as Time reported late on Tuesday, the company has recently dropped an important part of its flagship safety policy.—AO