Uber Says AI Vendor Breach Exposed US Drivers' License Data
Key Takeaways
- •Hackers used stolen internal credentials to access Uber's account on Fireworks AI's platform, exposing drivers' personal data including license details.
- •Disclosed figures show 7,870 affected people across South Carolina, Massachusetts, and Rhode Island, but the nationwide total remains unknown because California and Illinois have not released their counts.
- •Fireworks disclosed that account and billing information plus certain customers' secrets, datasets, and unfinished AI models were accessed, though payment card numbers were unaffected.
- •Fireworks linked the attacks to LLMJacking, a scheme that hijacks access to large language models and monetizes it by running AI workloads at the account holder's expense.
- •Uber's prompt vendor disconnection, driver notifications, and credit-monitoring offer contrast with its concealed 2016 breach, which resulted in a $148 million settlement in 2018.

Uber has begun notifying thousands of drivers across the United States about a security incident at one of its third-party artificial intelligence providers, which may have exposed drivers' personal data, including their license details. The company described the incident in notices sent to drivers dated November 10.
How Uber learned its vendor had been breached
According to the notice, the provider contacted Uber on October 9, 2026, after hackers obtained internal credentials and used them to access Uber's account on the vendor's platform. Uber said it cut off the data connection immediately and launched an investigation. By October 13, the company had confirmed that unauthorized parties had accessed drivers' personal data.
Affected drivers are being offered complimentary monitoring and fraud assistance through Cyberscout. Uber said no misuse of the exposed information has been reported so far.
The driver notice is dated November 10 — 28 days after the company verified the unauthorized access. That gap does not, however, indicate a violation of breach reporting requirements. In the United States, breach-notification rules are set state by state rather than by a single federal deadline, and many statutes — including California's — call for disclosure in the most expedient time possible and without unreasonable delay.
The provider is Fireworks AI
The affected vendor is Fireworks AI, a company that gives developers access to AI models, according to TechCrunch. Uber confirmed to the publication that Fireworks held some of its data.
Reported figures so far show 7,870 affected people across three states: 3,430 in South Carolina, 2,567 in Massachusetts, and 1,873 in Rhode Island. The South Carolina Department of Consumer Affairs independently confirmed its state's count.
The full extent of the breach remains unclear. California and Illinois have not released their figures, meaning the 7,870 total covers only the three states with disclosed counts. The nationwide number of affected drivers is still unknown.
A credential theft tied to "LLMJacking"
In its November 9 security bulletin, Fireworks announced that the incident was more serious than stolen credentials. Account and billing information were accessed, and certain customers had their secrets, datasets, and unfinished AI models accessed. Fireworks specified that payment card numbers were unaffected.
The company also disclosed a second breach. On October 23, a cloud services provider hosting Fireworks systems was itself victimized, and hackers used access to that cloud service the following day to obtain further customer information. Fireworks reported no additional successful breach after October 25.
Fireworks has connected the attacks to LLMJacking, a scheme based on stolen access to AI models. The term refers to hijacking access to large language models and monetizing it by running AI workloads at the account holder's expense. In a separate case, Google and Mandiant revealed that unauthorized AI use incurred costs of more than $100,000 within a single day — expenses not associated with Uber.
Why this breach is not the 2016 cover-up
Uber has faced a major data breach before, but its response this time was different. In 2018, California's attorney general announced a $148 million settlement over a concealed 2016 breach affecting 57 million people. Investigators said Uber paid hackers $100,000 to remain silent and failed to notify more than 174,000 California drivers.
In the current incident, Uber says it promptly disconnected the affected provider, notified drivers, and offered credit monitoring. Public disclosures so far show no evidence of a similar cover-up.
Third-party risk is now the bigger exposure
Uber's latest breach highlights a growing concern: companies can lose sensitive information through the vendors they depend on. The World Economic Forum found that 65% of large companies consider third-party and supply-chain vulnerabilities their biggest cyber-resilience challenge, up from 54% in 2025. Meanwhile, IBM estimated the global average cost of a breach at $4.99 million, and Verizon examined more than 22,000 confirmed breaches across 145 countries. These figures reflect broader cybersecurity risks, not Uber's actual losses.
Cryptopolitan previously reported a similar third-party AI-tool compromise involving Vercel.
For Uber, three questions remain: How many drivers were affected nationwide? What specific information was exposed? What safeguards controlled Fireworks' access to Uber data? The company's public notice leaves these questions unanswered.