Overview
A security incident involving a third-party technology provider has exposed personal information belonging to thousands of Uber drivers in the United States.
The incident became public on 13 November 2026, when technology reporting revealed that Uber was notifying drivers whose personal and driver's license information had been accessed through a compromised service-provider account.
The provider was identified by TechCrunch as Fireworks AI, a company that provides infrastructure for developers using open AI models. Uber confirmed to the publication that the incident involved Fireworks.ai, which had access to some Uber data.
The incident is significant because it demonstrates how a security compromise at one technology provider can expose data belonging to a much larger organization.
It also raises important questions for companies that use external AI platforms, model-hosting providers and cloud services in their internal systems.
What happened?
According to Uber's breach notification, a service provider informed the company on 9 October 2026 that an unauthorized third party had compromised internal credentials and used them to access Uber's account with that provider.
Uber says it immediately severed data connections with the provider and launched an investigation.
By 13 October, its investigation had determined that the compromised credentials had been used to access Uber data and that personal information belonging to affected recipients was included. The notification letter was dated 10 November.
The California Attorney General's website identifies Uber Technologies, Inc. and lists the known breach date as 7 October 2026. The notice and accompanying PDF provide the company's description of the subsequent discovery and response.
The public notification does not name the provider.
TechCrunch identified it as Fireworks AI and reported that an Uber spokesperson confirmed the vendor relationship. Fireworks also acknowledged that unauthorized access had affected customer data in its systems.
This was not described as an attack that directly compromised Uber's main internal systems. The publicly reported access path involved credentials at an external service provider.
How many drivers were affected?
Breach notifications reviewed by TechCrunch identified driver's license information belonging to residents in several US states.
| State | Reported affected individuals |
|---|---|
| Massachusetts | 2,567 |
| South Carolina | 3,430 |
| Rhode Island | 1,873 |
| Combined reported count | 7,870 |
These figures represent the reported counts for the three states, not a confirmed nationwide total.
Illinois filings indicated that some residents were affected, but the number was still pending. Uber also notified drivers in California without publishing a corresponding total in the reporting reviewed.
The number of affected people could therefore be higher.
Reuters independently reported the incident and confirmed that the California notification described unauthorized access through a compromised service-provider account. Its report also identified the Rhode Island figure of approximately 1,873 affected residents.
What was exposed?
Uber's notice says personal information was among the data accessed.
TechCrunch's review of state notices identified driver's license information as part of the exposed data.
The full scope of personal information for every affected individual was not publicly disclosed in the reporting reviewed for this article.
Uber said it had not identified any misuse of the exposed information at the time of its notification. It is providing affected people with credit monitoring and fraud assistance through Cyberscout, a TransUnion company. The notification provides a 90-day enrolment period from the date of the letter.
What Fireworks AI disclosed about its security incident
Fireworks AI published an initial security bulletin on 9 October, describing unauthorized use of internal credentials detected on 6 October.
The company initially said that an unauthorized party had accessed environment variables used in evaluation jobs in a limited number of customer accounts. It also said it had started containment and credential revocation.
On 9 November, Fireworks published a further update.
It said its investigation had identified additional impacts involving account and billing information. For a limited set of accounts, the company also found impacts to secrets, datasets and incomplete model weights.
Fireworks said it had worked with Palo Alto Networks Unit 42 on forensic analysis and security hardening. It described measures including credential replacement, reduced privileges, network isolation and stronger logging.
The company also said the evidence indicated links to a financially motivated threat actor interested in selling unauthorized large-language-model access tokens, a practice commonly described as LLMjacking.
That is Fireworks' assessment of its broader security incident. The public information does not establish that a named criminal group was responsible for the specific Uber data access.
Fireworks also described a separate incident involving a third-party cloud provider that hosted its research and development machines on 23–24 October. The company said that event led to additional customer-data impact.
Public disclosures do not establish that this later incident was the route through which Uber's data was accessed. The two timelines should not be conflated.
Why this is important for AI and SaaS security
Modern software companies increasingly depend on external providers for:
- AI model hosting.
- Inference infrastructure.
- Data processing.
- Application monitoring.
- Analytics.
- Cloud storage.
- Customer-support automation.
- Development and evaluation environments.
These integrations can provide substantial efficiency and reduce infrastructure costs.
However, every connection can also create another path through which credentials or data may be accessed.
A company may have strong internal authentication and network controls while still exposing information through a provider account that has broader permissions than necessary.
The Uber incident illustrates why vendor risk management must include more than reviewing a provider's security certifications.
What SaaS companies should learn
Technology companies should consider several practical controls.
1. Apply least-privilege access
Third-party vendors should receive only the information and permissions necessary to perform their specific function.
2. Separate production and development credentials
A key used for testing or evaluation should not automatically provide access to customer or production information.
3. Use short-lived and scoped credentials
Long-lived tokens increase the period during which compromised credentials can be abused.
4. Monitor vendor access
Companies should retain logs showing which external services accessed sensitive data, when they accessed it and what permissions were used.
5. Maintain an incident response plan for suppliers
Contracts and technical integrations should support prompt notification, credential revocation, access isolation and forensic investigation.
6. Minimize sensitive information shared with AI providers
Data minimization can reduce the consequences of a breach, even when an external provider is compromised.
What remains uncertain?
Several important details remain unavailable publicly.
Uber has not disclosed a total number of affected drivers across all states.
The state notification sample contains placeholders for some personal-data elements, so the exact information exposed may differ among recipients.
Fireworks has published information about its own incident and remediation, but the public record does not provide a complete technical reconstruction of the access path to every Uber record.
No named threat actor has been confirmed as responsible for the Uber data access.
The absence of identified misuse is also not proof that exposed information can never be used fraudulently.
What to watch next
The next important developments will be additional state breach notices, further disclosure from Uber and Fireworks, and any updated forensic findings.
Technology companies using external AI infrastructure should review their own vendor permissions, secrets-management processes and data-access logs.
The larger lesson is that AI supply-chain security is not limited to protecting models and prompts. It also includes protecting every credential, dataset and enterprise account connected to the providers that make AI services possible.
Why this is trending
- TechCrunch: Published the detailed report on 13 November, identifying Fireworks AI as the vendor and reporting state-level driver's license exposure figures.
- Reuters: Independently reported Uber's notification, the compromised service-provider credentials, the investigation timeline and the Rhode Island affected count.
- California Attorney General: Published Uber's breach notification sample, identifying the organization and breach date. The attached notice describes the October 9 notification, October 13 investigation finding and remediation steps.
- Fireworks AI: Published its initial incident bulletin and a 9 November update describing additional findings and security hardening.
- Hacker News: No matching current discussion about the Uber/Fireworks incident was verified on the captured front page.
- Reddit: No substantive current discussion with verifiable engagement data was identified.
- daily.dev: No matching current report was verified.
- X: Direct search did not provide verifiable current posts or engagement figures.
The strongest evidence comes from official breach notices, Fireworks' security bulletins and independent reporting. No unsupported claim of widespread social-media traction is made.
Sources
- California Attorney General — Uber Technologies, Inc. Breach Notification — November 2026. Official state filing identifying Uber and the reported breach date.
- Uber — Multistate Security Incident Notification — Notice dated 10 November 2026. Primary notification describing compromised provider credentials, the investigation timeline and customer assistance.
- TechCrunch — Uber confirms hackers stole thousands of drivers' license information in data breach — 13 November 2026. Independent reporting identifying Fireworks AI and the reported state-level figures.
- Reuters — Uber says US drivers' license data stolen in data breach — 13 November 2026. Independent confirmation of the notification and compromised-provider access path.
- Fireworks AI — October 2026 Security Incident — 10 October 2026. Provider's initial incident description, credential exposure and containment.
- Fireworks AI — Update on October 2026 Security Incident — 9 November 2026. Further findings, affected data categories, forensic assistance and security-hardening actions.
Further reading
- California Attorney General — Uber Technologies, Inc. Breach Notification — November 2026. Official state filing identifying Uber and the reported breach date.
- Uber — Multistate Security Incident Notification — Notice dated 10 November 2026. Primary notification describing compromised provider credentials, the investigation timeline and customer assistance.
- TechCrunch — Uber confirms hackers stole thousands of drivers' license information in data breach — 13 November 2026. Independent reporting identifying Fireworks AI and the reported state-level figures.
- Reuters — Uber says US drivers' license data stolen in data breach — 13 November 2026. Independent confirmation of the notification and compromised-provider access path.
- Fireworks AI — October 2026 Security Incident — 10 October 2026. Provider's initial incident description, credential exposure and containment.
- Fireworks AI — Update on October 2026 Security Incident — 9 November 2026. Further findings, affected data categories, forensic assistance and security-hardening actions.
Building a product in this space?
MORE FROM THE NEWSROOM
Related news
CybersecurityMalicious Optimizely Scripts Reached 178 Organizations, Exposing Website Visitors to Data Theft
A malicious JavaScript campaign distributed through Optimizely Web reached visitors across 178 organizations, according to security researcher Scott Helme. The payload included cryptocurrency address substitution, payment-card skimming, cookie theft and iOS exploit delivery mechanisms, highlighting the security risks associated with trusted third-party website scripts.
CybersecuritySignDrop Windows Zero-Day Exposes Local SYSTEM Privilege-Escalation Path; Microsoft Investigates
Security researcher Abdelhamid Naceri has disclosed SignDrop, a Windows privilege-escalation vulnerability involving the Common Log File System and Microsoft Defender. A public proof of concept demonstrates a path from an existing local user account to SYSTEM privileges, although Microsoft says Defender detects and blocks the associated exploit behavior and that it has not observed exploitation in the wild.
CybersecurityWFX-Loot Campaign Uses Hijacked GitHub Accounts to Steal CI/CD Secrets From Thousands of Repositories
Security researchers have identified a campaign called WFX-Loot that used stolen GitHub credentials to distribute a credential-stealing payload through GitHub repositories. The attack targeted CI/CD workflows, collected cloud and AI-service credentials, and attempted to maintain access even after affected organizations rotated their secrets.
