OpenAI and the TanStack npm attack (May 2026): what OpenAI confirmed
In May 2026 OpenAI said the TanStack npm supply-chain attack reached two of its employees’ devices. OpenAI holds SOC 2 Type 2 and ISO 27001 for its business products.
Short answer
Was OpenAI hacked?
In a limited way, yes. OpenAI said on 13 May 2026 that two employee devices were compromised by the TanStack npm supply-chain attack, leading to limited credential exfiltration involving a small subset of internal source-code repositories.
OpenAI said it found no evidence that user data, production systems or intellectual property were compromised, and that the two devices lacked updated configurations that would have blocked the malicious download because those controls were still being rolled out.
Timeline at a glance
13 May 2026OpenAI: two employee devices hit; limited credential exfiltration; no user data. (OpenAI)Confirmed
Last updated October 31, 2026. Every entry is dated and linked to its source; claims by attackers are labelled as claims, never stated as fact.
OpenAI’s security posture at the time
What OpenAI publicly said about its security certifications and controls, how big it is, and how its shares moved, each tied to a source. A certificate covers a stated scope at a point in time; it is not a guarantee about every system. Reviewed 31 October 2026. Compare all 2026 breaches.
Security certifications publicly held
SOC 2SOC 2 Type 2Scope: ChatGPT Enterprise, ChatGPT Edu and the API (latest report noted on the portal covered Jan–Jun 2025).Source: OpenAI Trust Portal
ISO 27001ISO/IEC 27001:2022 and ISO/IEC 27701:2019Scope: The information security and privacy management systems supporting the OpenAI API, ChatGPT Enterprise and ChatGPT Edu.Source: OpenAI — Security and privacy
Scope vs breach. The published scope does not say whether the breached system was covered. The certificates name customer-facing services; whether employee endpoints and internal repositories sit inside the ISMS scope is not published.
Publicly known security controls and platforms
OpenAI says the two affected employee devices lacked updated configurations that would have blocked the malicious package download, because those controls were still being rolled out. (company statement; OpenAI — Our response to the TanStack npm supply-chain attack)
Size
Private; no audited revenue published
Listing: Private
Share price around the disclosure
Not listed.
OpenAI incident timeline, in detail
Confirmed = acknowledged by OpenAI, a regulator filing or a court. Reported = credible reporting OpenAI has not fully confirmed. Unverified claim = an attacker or leak-site claim.
13 May 2026ConfirmedSoftware supply chain
Poisoned npm package reaches two OpenAI employee devices
What happened. OpenAI said two employee devices were compromised by the TanStack npm supply-chain attack, leading to limited credential exfiltration involving a small subset of internal source-code repositories. It said it found no evidence that user data, production systems or intellectual property were compromised.
Impact. OpenAI rotated code-signing certificates and asked macOS users to update desktop apps, per reporting. People or records affected: No user data affected, per OpenAI.
OpenAI publicly held SOC 2 and ISO 27001. Stated scope: ChatGPT Enterprise, ChatGPT Edu and the API (latest report noted on the portal covered Jan–Jun 2025). The information security and privacy management systems supporting the OpenAI API, ChatGPT Enterprise and ChatGPT Edu. The published scope does not say whether the breached system was covered.
The weak link was in the software supply chain
The entry point was a dependency, token or third-party software rather than the company’s own front door. Certification scopes usually cover the company’s management system; the integrity of every package, token and developer workstation is a control that has to be checked separately.
What it means for OpenAI OpenAI customers and developers — and what to do
What OpenAI’s customers, partners and security teams can do now, based only on what has been disclosed.
1.Check whether you used the same component
If the incident involved a named package, token or product, check your own inventory for the same component and versions.
2.Rotate shared secrets
Rotate any credentials, tokens or keys shared with the affected company until it confirms they were not exposed.
3.Ask the supplier for its findings
Ask what was taken, whether code-signing or update channels were affected, and what you need to update.
In a limited way, yes. OpenAI said on 13 May 2026 that two employee devices were compromised by the TanStack npm supply-chain attack, leading to limited credential exfiltration involving a small subset of internal source-code repositories.
Did OpenAI have SOC 2 or ISO 27001 certification?+
OpenAI publicly held SOC 2 and ISO 27001. Stated scope: ChatGPT Enterprise, ChatGPT Edu and the API (latest report noted on the portal covered Jan–Jun 2025). The information security and privacy management systems supporting the OpenAI API, ChatGPT Enterprise and ChatGPT Edu. The published scope does not say whether the breached system was covered.
More in the Breach Index
Same sourced, verification-tagged timeline, company by company — compare how peers have fared.
Independent analysis for security-leadership education; not affiliated with or endorsed by OpenAI. Facts are attributed to the dated sources above, and anything disputed or unconfirmed is labelled as such. See the full Breach Index.
🎉 New here? Use code WELCOME10 for 10% off any plan at checkout