OpenAI stated it has no energetic plans to make use of Google’s in-house chip to energy its merchandise, two days after Reuters and different information retailers reported on the AI lab’s transfer to show to its competitor’s synthetic intelligence chips to satisfy rising demand.

A spokesperson for OpenAI stated on Sunday that whereas the AI lab is in early testing with a few of Google’s tensor processing models (TPUs), it has no plans to deploy them at scale proper now.

Google declined to remark.

Whereas it is not uncommon for AI labs to check out completely different chips, utilizing new {hardware} at scale might take for much longer and would require completely different structure and software program assist. OpenAI is actively utilizing Nvidia’s graphics processing models (GPUs), and AMD’s AI chips to energy its rising demand. OpenAI can also be growing its chip, an effort that’s on observe to satisfy the “tape-out” milestone this yr, the place the chip’s design is finalized and despatched for manufacturing.

OpenAI has signed up for Google Cloud service to satisfy its rising wants for computing capability, Reuters had solely reported earlier this month, marking a shocking collaboration between two distinguished rivals within the AI sector. Many of the computing energy utilized by OpenAI can be from GPU servers powered by the so-called neocloud firm CoreWeave.

Google has been increasing the exterior availability of its in-house AI chips, or TPUs, which have been traditionally reserved for inner use. That helped Google win prospects, together with Large Tech participant Apple, in addition to startups like Anthropic and Secure Superintelligence, two ChatGPT-maker rivals launched by former OpenAI leaders.

© Thomson Reuters 2025