Uber is notifying 149,346 drivers that their names, addresses, dates of birth and driver's license numbers were compromised during an October data breach at Fireworks AI, one of its technology vendors. The company filed notices this week with attorneys general in California, Illinois, Massachusetts, Oregon, Rhode Island, South Carolina, Texas and Vermont.

Uber learned on October 9 that hackers had stolen internal credentials from Fireworks AI and used them to access Uber's data stored in the vendor's environment. The breach extended beyond Uber. Fireworks AI, which provides access to open-weight AI models for developers, also exposed customer data at other clients, including Notion and Cursor, the coding startup. Fireworks attributed the incident to a financially motivated cybercriminal group targeting LLM tokens for resale; no group has claimed responsibility.

Uber had alerts configured to detect unauthorized access to its Fireworks environment, but they failed to trigger because attackers used Fireworks' own internal credentials, bypassing external breach detection. A Fireworks spokesperson deferred additional questions to the vendor.

This is Uber's second known data breach this year. In a separate incident, an extortion group claimed responsibility for data theft involving Uber Freight, the company's truck-shipping unit.

The repeated incidents expose a structural vulnerability in how large platforms manage third-party AI infrastructure. As startups like Fireworks become operational dependencies—not just vendor services—their security posture directly determines customer risk. Uber's breach reveals a gap: internal credential theft at a vendor can circumvent a customer's own detection systems. For platforms managing driver data at scale, that gap is material.