Uber has joined mobility leaders Lucid and JLR in supporting NVIDIA’s newly released Alpamayo family of open-source AI models, simulation tools and datasets. The company highlighted Alpamayo 1’s 10 billion-parameter chain-of-thought vision language action model, the AlpaSim high-fidelity simulation framework and more than 1,700 hours of driving data in the Physical AI Open Datasets as critical to fast-tracking safe, reasoning-based level 4 deployment. Uber’s global head of autonomous mobility and delivery noted that handling unpredictable, long-tail driving scenarios remains a defining challenge for autonomy and praised Alpamayo’s potential to improve transparency and accelerate commercial rollout of self-driving services.

Research firm Melius Ratings has downgraded Uber from Hold to Sell, citing growing risk from autonomous vehicle entrants. Analysts pointed out that as Uber positions itself as a demand aggregator for emerging AV players through strategic partnerships, it also becomes exposed to competitive threats that could erode returns. The downgrade followed an intraday share decline of over 1%, and Melius warned that, if companies like Waymo or Tesla expand their standalone operations in the U.S., downside risks may materialize that are not currently reflected in Uber’s valuation.