# Fireworks AI discloses second breach: billing data, datasets and incomplete model weights accessed

*Security incident · 10 Nov 2026*

Fireworks AI, an inference platform for open-source models such as Llama, Mistral and Gemma, expanded its disclosure of a security incident on 9 November 2026.

**What happened.** In early October the company reported unauthorized use of internal credentials. Attackers had accessed environment variables (API keys and tokens) in a limited number of accounts; Fireworks linked this to "LLMjacking", i.e. selling stolen LLM access. On 23 October a third-party cloud provider hosting Fireworks R&D machines was breached. On 24 October the attacker used that route to reach Fireworks development machines and ran an unauthorized workload.

**What was accessed.** According to Fireworks: resource metadata (identifiers, names, descriptions, creation times), billing information (billing and shipping addresses, invoice and payment history, account balance, credit usage) and, for some customers, secrets, datasets and incomplete model weights. Payment card numbers were not affected because Fireworks does not store them.

**What is not affected so far.** Fireworks says it found no impact on Inference, the Training SDK or Nexus, which are zero data retention by default. It stresses this is limited to the investigation to date. As of 25 October it saw no further successful access by the attacker.

**Response.** Credentials and service accounts were replaced, network paths restricted and affected systems isolated. Unit 42 (Palo Alto Networks) is investigating, and law enforcement has been notified. Self-serve dataset uploads and downloads for Batch Inference and dataset-based training are disabled for now.

**Why it matters for users of LLM APIs.** Billing data and datasets stored with a provider can be exposed even when prompts are not retained. Rotate API keys, check what you uploaded, and prefer zero-retention settings. Fireworks is not listed as a product in our database.

## Sources

- [BleepingComputer: Fireworks AI discloses new breach after cloud provider hack](https://www.bleepingcomputer.com/news/security/fireworks-ai-discloses-new-breach-after-cloud-provider-hack/)
- [Unite.AI: Fireworks AI Expands Breach Disclosure to Datasets and Model Weights](https://www.unite.ai/fireworks-ai-expands-breach-disclosure-to-datasets-and-model-weights/)
- [Fireworks AI security bulletin (update)](https://fireworks.ai/security/bulletins/october-2026-security-incident-update)

---

- **Canonical URL:** https://cryptollmss.com/en/news/fireworks-ai-second-breach-billing-data-datasets-model-weights/
- **As of:** 10 Nov 2026
- **Data licence:** CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)

Please cite: Source: cryptollmss.com (cryptovss), as of 10 Nov 2026, https://cryptollmss.com/en/news/fireworks-ai-second-breach-billing-data-datasets-model-weights/
