The DOJ has sent subpoenas to NVIDIA and other companies as part of an antitrust probe, as reported by Bloomberg. The federal government is seeking evidence that the company violated antitrust laws with regard to its AI processors. The presence of these subpoenas means the DOJ is one step closer to launching a formal complaint.
Officials speculate that NVIDIA is making it difficult for other companies to switch hardware suppliers and that it “penalizes buyers that don’t exclusively use its artificial intelligence chips.” This probe started in June, but recently escalated to include legally binding requests for information.
At the root of the DOJ probe is NVIDIA’s recent acquisition of RunAI, a company that makes software for managing AI computing tasks. The concern is that this purchase will make it harder for business customers to switch away from NVIDIA chips, as it would also necessitate a change in software.
However, that’s not the only reason behind this investigation. Regulators are also looking into whether NVIDIA gives preferential treatment to customers who exclusively use its technology or buy its complete systems. This special treatment allegedly includes first dibs on hardware and related supplies and unique pricing models.
NVIDIA has offered a terse response, telling Bloomberg that it “wins on merit, as reflected in our benchmark results and value to customers, who can choose whatever solution is best for them.” The inference here is that the company’s market dominance comes down to hard work and not sweetheart deals.
The investigation is still in its early days, as it hasn’t yet blossomed into a formal complaint. The company’s stock took a hit ahead of the DOJ announcement, but that was likely due to continuing delays for its Blackwell AI chip. However, the stock is still up more than double this year as the AI boom continues to do its thing.
This article originally appeared on Engadget at https://www.engadget.com/ai/doj-subpoenas-nvidia-as-part-of-antitrust-probe-regarding-ai-processors-153435877.html?src=rss
