Anthropic acquires Opaque Systems (confidential and verifiable AI)
Confirmed
The takeaway
On Oct 13, 2026 Opaque Systems, a UC Berkeley RISELab spin-out that built confidential-computing and “verifiable AI” tools (cryptographic evidence of what an AI system ran, where and under which policies), said its team is joining Anthropic and the company is winding down.
Status
- Claim
Confirmed
- Our reporting
- High confidence
- Importance
- 3 of 5
- Last verified
- 19 October 2026
Your AI and this story
- GPT-6 Astra166 days after its cutoff
- Claude Opus 5.5105 days after its cutoff
- Gemini 4 Argon196 days after its cutoff
- Grok 4.7135 days after its cutoff
None of these four assistants can know about it. The closest, Claude Opus 5.5, stops 105 days before it.
Key facts
- Opaque’s statement (opaque.co, post ‘OPAQUE’s Next Chapter’, Oct 13): ‘Our team is joining Anthropic to carry our work forward, and OPAQUE Systems is winding down its operations.’ The post has no individual byline
- Terms: not disclosed (SecurityWeek, Oct 19). We found no Anthropic announcement of the deal; anthropic.com/news lists none as of Oct 19
- What Opaque did: started about a decade ago as a UC Berkeley RISELab project on keeping data private while in use (confidential computing), then moved into ‘verifiable AI’. Open-source work it lists: AgenTrust, Agent Governance Toolkit, TRACE, Confidential MCP and Weight Custody Manifest (Opaque; SecurityWeek)
- People (SecurityWeek): co-founders Ion Stoica (Databricks and Anyscale co-founder) and Raluca Ada Popa (now at Google DeepMind); CEO Aaron Fulkerson since October 2023
- Funding (SecurityWeek): a $24M Series B this year at a $300M valuation; customers it listed included Anthropic, ServiceNow and Accenture
- Open question: neither Opaque nor SecurityWeek says what happens to existing customers or to the open-source projects
What happened
Opaque Systems posted on Oct 13, 2026 that its team “is joining Anthropic to carry our work forward” and that the company is winding down. SecurityWeek covered it on Oct 19 as an acquisition by Anthropic with undisclosed terms. Opaque grew out of UC Berkeley’s RISELab (co-founded by Ion Stoica and Raluca Ada Popa) and sold confidential-computing tools that let AI systems work on sensitive data while hardware keeps it protected. It later turned to “verifiable AI”: cryptographic evidence of what an AI system ran, where and under which policies, with open-source projects for agent governance, confidential MCP and model weight custody. It had raised a $24M Series B at a $300M valuation this year and counted Anthropic among its customers.
Why it matters
It is a small team acquisition, but its subject matters to Anthropic: protecting model weights and customer data, and proving to customers and governments what an agent actually did. Anthropic has not explained the deal itself, so how it will use the team is not known.
Sources
3 sources from 2 sites. Numbers match the chips in the text.
3 sources: 2 primary, 1 press
Primary
- Opaque: From Berkeley research to the future of AI, OPAQUE’s next chapter (Oct 13)opaque.co, official
- Opaque home page: team joining Anthropic, company winding downopaque.co, official
Press
- SecurityWeek: Anthropic Acquires Confidential AI Startup Opaque Systems (Oct 19)securityweek.com, press
Changes
- Filed (SecurityWeek, Opaque’s post)