← All stories
● Covered by 5 sources · 6 reportsMedium impact3 negative3 neutral

Fireworks AI discloses new breach after cloud provider hack, impacting customer data

🔄 Updated 5d ago — new reporting from SecurityWeek, TechCrunch
New to BrevFeed? We gather this story from every outlet covering it into one summary — ranked by real-world impact, not just the latest headline — so you never miss what matters. What is BrevFeed? →

Key points

  • Fireworks AI experienced a new security breach on October 23.
  • Attackers accessed development machines via a compromised third-party cloud provider.
  • Customer resource metadata, billing, and incomplete model data were exposed.
  • This follows an earlier breach in October involving API keys and tokens.
  • Fireworks AI does not store customer payment card numbers.
  • The initial breach was detected on October 6, 2026, at 9 a.m. PDT.
  • Fireworks AI published a notice about the initial breach on October 10, 2026.
  • The breach also affected Notion AI chats.
  • Fireworks AI hired Palo Alto Networks Unit 42 to investigate the breach.
  • A financially motivated cybercriminal group was responsible for the attacks.
  • Fireworks AI was founded in 2022 and has over 10,000 customers.
  • Uber drivers' personal and driver's license information was stolen.
  • Uber was notified on October 9 by Fireworks AI about the breach.
  • The stolen data includes driver's licenses for thousands of individuals across multiple US states.
  • The supply chain attack on Cursor was discovered on October 29.
  • Attackers made two malicious versions of the anthropic.claude-code extension (2.1.312 and 2.1.313) available.
  • The malicious Claude Code extensions were available for roughly 11 hours.
  • Anthropic was never hacked; all malicious activity occurred in Cursor’s environment.
  • A Cursor access key issued to Fireworks was compromised.
  • The compromised Cursor access key allowed attackers to copy data from a read-only storage bucket.
  • Uber is notifying close to 150,000 drivers about the data theft.
  • Uber filed notices with several U.S. attorneys general, including California and Iowa.

New Breach Details Emerge

Fireworks AI, a startup providing high-performance inference platforms for AI models, has disclosed a second security breach. This incident, which occurred on October 23, involved attackers gaining access to the company's development machines after compromising a third-party cloud provider that hosted these systems. The company detected the activity on the same day and took immediate remediation measures.

Data Compromised in the Attack

During this breach, the threat actors accessed various types of customer data. This included resource metadata such as Fireworks resource identifiers, display names, descriptions, and creation times. Billing information, including billing and shipping addresses, invoice and payment history, account balance, and credit usage, was also exposed. Additionally, incomplete model weights and datasets for some customers were accessed. Fireworks confirmed that customer payment card numbers were not compromised as they are not stored by the company.

Previous Security Incident

This latest breach follows an earlier security incident in early October. In that event, attackers accessed environment variables containing API keys and tokens stored as secrets for a limited number of accounts. Fireworks linked this earlier attack to a financially motivated cybercriminal group aiming to sell unauthorized access to Large Language Models (LLMs), a practice known as 'LLMJacking'.

Response and Remediation

Following the October 23 breach, Fireworks AI revoked and replaced compromised credentials and service accounts. The company also restricted unauthorized network pathways, isolated affected systems, and tightened access controls to data storage and analytics systems to prevent further unauthorized access. As of October 25, Fireworks had not detected further evidence of successful unauthorized access by the threat actor on its customer-facing platform.

Updates

🕒 2026-11-16 · new reporting from SecurityWeek, TechCrunch
  • The supply chain attack on Cursor was discovered on October 29.
  • Attackers made two malicious versions of the anthropic.claude-code extension (2.1.312 and 2.1.313) available.
  • The malicious Claude Code extensions were available for roughly 11 hours.
  • Anthropic was never hacked; all malicious activity occurred in Cursor’s environment.
  • A Cursor access key issued to Fireworks was compromised.
  • The compromised Cursor access key allowed attackers to copy data from a read-only storage bucket.
  • Uber is notifying close to 150,000 drivers about the data theft.
  • Uber filed notices with several U.S. attorneys general, including California and Iowa.
🕒 2026-11-13 · new reporting from TechCrunch
  • Uber drivers' personal and driver's license information was stolen.
  • Uber was notified on October 9 by Fireworks AI about the breach.
  • The stolen data includes driver's licenses for thousands of individuals across multiple US states.
🕒 2026-11-10 · new reporting from The Hacker News, The Record
  • The initial breach was detected on October 6, 2026, at 9 a.m. PDT.
  • Fireworks AI published a notice about the initial breach on October 10, 2026.
  • The breach also affected Notion AI chats.
  • Fireworks AI hired Palo Alto Networks Unit 42 to investigate the breach.
  • A financially motivated cybercriminal group was responsible for the attacks.
  • Fireworks AI was founded in 2022 and has over 10,000 customers.

✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →

The daily brief

One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.

One email a day. Unsubscribe in one click, any time.

Today's brief

Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.

~4 min · 3 stories · Nov 21

▶ Play today's brief Listen on Spotify

New every morning, and the back catalogue is archived by date.

How outlets covered it

Uber is notifying nearly 150,000 drivers that their personal information, including driver's license numbers, was stolen during a data breach at its third-party AI provider, Fireworks AI. The hackers used Fireworks' internal credentials to access Uber's data, bypassing Uber's security alerts, and the incident highlights supply chain risks in data management.

A supply chain attack on Cursor's VS Code Extension Marketplace, which distributed malicious versions of the Claude Code extension, stemmed from compromised credentials at Fireworks AI. Attackers used an access key stolen from Fireworks AI to gain access to Cursor's internal systems and inject malware, affecting up to 5,000 users.

Uber is notifying thousands of drivers that their personal and driver's license information was stolen in an October data breach at Fireworks AI, a tech provider. Hackers accessed Uber's data through stolen internal credentials from Fireworks AI, which also affected other Fireworks AI customers like Notion. The stolen data includes driver's licenses for thousands of individuals across multiple US states.

AI infrastructure company Fireworks experienced two data breaches in October, resulting in the theft of customer billing information, account data, and AI-specific datasets. A financially motivated cybercriminal group was responsible for the attacks, which also impacted customers of Notion, a digital workspace company.

Fireworks AI revealed that a security incident, initially reported in early October, was more extensive than first believed, impacting customer datasets, incomplete model weights, and billing information for a limited number of accounts. The breach originated from unauthorized access to its development environment via a third-party cloud provider, leading to further data compromise.

Fireworks AI, an AI infrastructure startup, disclosed a new security breach where attackers accessed its development machines by compromising a third-party cloud provider. This incident, occurring on October 23, led to the exposure of customer resource metadata, billing information, and incomplete model weights and datasets. This follows an earlier breach in early October where API keys and tokens were accessed.