Overview
Severity: CRITICAL | Affected: Cognition Corp | Category: breach
AI startup Cognition Corp, known for its advanced reasoning models, confirmed a significant security breach today. Attackers exploited a misconfigured cloud service to gain access to the company's core infrastructure, exfiltrating the proprietary model weights for their upcoming 'Synapse-2' model. Additionally, a database containing over 10 million user prompt-and-response pairs from their enterprise API was compromised. The breach poses a critical threat, as the stolen weights could allow malicious actors to replicate the model, while the user data leak raises severe privacy concerns. The company has engaged a cybersecurity firm to investigate the incident and has begun notifying affected customers. This event highlights the growing targeting of AI companies for their valuable intellectual property and data.