Braintrust Probes Cloud Security Incident, Urges Customers to Rotate API Keys

New updates have been reported about Braintrust.

Braintrust is investigating a security incident in one of its Amazon Web Services environments that exposed customer API keys used to access AI models, prompting the startup to ask all clients to revoke and replace any keys stored on its platform. The company said it has so far identified one affected customer and no evidence of broader compromise, but moved to lock down the impacted account, tighten access controls across related systems, and rotate its own internal secrets.

Management describes the notification to all customers as a precautionary step while the root cause of the breach remains under review, underscoring the potential downstream risk if stolen API keys were used to impersonate legitimate users in customer environments. The incident comes shortly after Braintrust’s $80 million Series B funding at an $800 million valuation and highlights the security expectations on its AI monitoring “operating system” for enterprise engineers, particularly for AI vendors and enterprises that integrate deeply with its service. Executives and security teams using Braintrust should immediately complete key rotation, review logs for anomalous activity, and reassess third-party cloud security posture given the growing trend of attackers targeting SaaS and cloud providers as leverage points into customer systems.