Prediction market data from Manifold Markets as of June 29, 2026 shows a 93% consensus that OpenAI will design and...

Prediction market data from Manifold Markets as of June 29, 2026 shows a 93% consensus that OpenAI will design and manufacture a custom AI chip by 2030 (source). This indicates strong market belief in OpenAI's vertical integration into chip design, a move that would reduce reliance on GPU suppliers like NVIDIA and potentially increase competition in the AI accelerator market. Additionally, there is a 24.57% probability that OpenAI will publicly share software to run its AI on chips from different providers (source), suggesting the company may also pursue a multi-provider compute strategy. While these are speculative probabilities, they reflect growing conviction in OpenAI's hardware ambitions. If realized, a custom chip would disrupt the existing GPU supply chain, benefiting foundries like TSMC while pressuring NVIDIA's dominance. The data points are early and based on betting markets, thus confidence is low, but the directional implication for AI infrastructure is clear: OpenAI is likely to invest in custom silicon.