top of page
Insights
Practitioner thinking, from the people doing the work.
We publish what we learn, the same instinct that leads us to ship our source code. No gated downloads, no thought leadership theatre. Writing on agentic AI governance, the EU AI Act and identity as the control plane.
Search for Midships Knowledge Base
! No Results found
The OpenAI and Hugging Face incident is unnerving. But perhaps not for the reason you think.
The recent OpenAI incident, where an AI agent escaped its intended environment and ultimately compromised Hugging Face infrastructure during a cyber capability evaluation, has generated plenty of discussion. It should. Any story involving autonomous AI, zero day vulnerabilities and production systems is naturally unsettling. However, I think there is a danger in drawing the wrong conclusion. Most organisations are not deploying AI in this way. Banks, insurers, governments and
Ajit Gupta
2 days ago3 min read
The governance gap Anthropic just exposed.
Anthropic just released ten finance agent templates. They cover the most expensive and most regulated parts of bank operations: KYC screening, month-end close, general ledger reconciliation, statement audit, valuation review, earnings review, and model building. The published benchmark score is 64.37 percent on Vals AI's Finance Agent benchmark. Anthropic is clear that this is industry-leading. It is also the best on offer. Roughly one in three finance tasks still fails. Read
Ajit Gupta
May 82 min read
Common Questions
Insights – FAQs
What does Midships publish?
Practitioner writing on agentic AI governance, the EU AI Act and identity as the control plane, from the people doing the work.
Are the insights gated?
No. There are no gated downloads and no thought leadership theatre.
bottom of page