Cybersecurity — 2026-04-20 Vercel confirms Context.ai-linked breach exposed customer environment variables; ShinyHunters lists $2M sale on BreachForums Vercel confirmed on 19 April that a compromise of Context.ai, a third-party AI assistant used by one of its employees, allowed the attacker to take over that employee's Google Workspace account and access some Vercel environments and environment variables not marked sensitive. A BreachForums post attributed to ShinyHunters offered the stolen data for $2 million. Vercel said it is working with Mandiant, peer firms, and law enforcement, and that values marked sensitive remain unread. Context.ai disclosed in parallel that it had identified and blocked unauthorized access to its AWS environment in March and that the attacker likely compromised OAuth tokens for some consumer users. The OAuth pivot through Context.ai into Vercel is very likely the most consequential AI-supply-chain incident to date. Vercel's Next.js deployment footprint spans fintech and crypto front-ends whose API keys sat in the exposed environment variables, and downstream secondary breaches are very likely to materialize within the next 30-60 days. ShinyHunters' $2M public listing almost certainly reflects sellable but non-exclusive data, credential and non-sensitive env-var exposure rather than structured PII, at a price calibrated for bulk buyers, not a single premium acquirer. Corroborated across Vercel's authoritative bulletin and three major outlets, employee-granted AI-tool OAuth scopes have become the dominant lateral-movement vector into SaaS tenancies, and detection tooling has not kept pace with the exposure surface. Sources: The Hacker News · TechCrunch · SecurityWeek · Vercel Knowledge Base