At a glance
| Client | Maven Capital Partners |
|---|---|
| Sector | Private equity and portfolio management |
| Service | AI Training for SMEs and the Public Sector |
| Delivered | Two one hour AI governance sessions, November 2025 and April 2026 |
| Result | Portfolio directors left with governance questions they could put to portfolio companies, supplied in a follow up PDF |
Why did Maven Capital Partners need AI governance training?
Maven Capital Partners told us it manages more than 300 portfolio companies. Those companies were already using or trying AI in different ways. That left the portfolio team with a practical question about oversight.
The senior team wanted a clearer picture of the technology. They needed plain English answers on what AI can do, where the risks sit, and what portfolio directors should be asking.
The work started as a one hour talk at a portfolio away day. It grew into a second workshop for a wider staff group, senior leaders included.
What did Insightful AI deliver?
The first session explained the technology and where it goes wrong. The November talk set out what AI, machine learning, deep learning, generative AI and agentic AI mean. Agentic AI is the kind that takes actions on your behalf rather than just answering questions. The session also covered UK adoption trends, shadow AI, data privacy, compliance, hallucinations and bias. Shadow AI is staff using tools the organisation has never approved. A hallucination is an answer the tool gives confidently that turns out to be wrong.
The second session focused on practical oversight. The April workshop explained how large language models work. These are the kind of AI behind tools like ChatGPT. It covered why they can sound certain and still be wrong, and why a person still has to make the call. It also covered sycophancy, where a tool agrees with you rather than correcting you, and prompt injection, where hidden text tricks a tool into following instructions it should ignore.
The workshop linked AI use to duties Maven already carries. The session covered UK GDPR, the Data Protection Act 2018 and the Data Use and Access Act 2025. It covered what the ICO expects and what the FCA has prioritised. It also covered SM&CR, the regime that makes named senior people accountable, and Consumer Duty, the rule requiring firms to deliver good outcomes for customers. Each was turned into a question Maven could put to itself and to its portfolio companies.
A follow up PDF gave Maven something to use. After the April workshop, Insightful AI sent a PDF setting out the main governance points, questions for portfolio companies, and where to look in the rules.
How does the governance work?
The training was risk led and evidence first. Insightful AI did not ask Maven to buy or adopt any AI tool or platform, and had no commercial interest in what Maven chose.
The controls covered in the training were:
- finding out which AI tools staff actually use
- checking what data those tools can reach
- reviewing default permissions
- writing down who oversees decisions about people, customers or money
- challenging what a tool produces before acting on it
- recording how AI use is governed and watched
This work spans two stages of the Insightful AI Adoption Roadmap: Build AI fluency and Establish governance. Against People, Process and Principles it weighs to People, because the training built the judgement that governance rests on. Principles were covered through the regulatory questions, and Process was left for Maven to design.
What were the results?
| Measure | Before | After |
|---|---|---|
| Senior team understanding | Maven wanted a clearer picture of AI and what it meant for governance | Ten members of the senior portfolio team attended a one hour briefing |
| Reach across staff | Training had reached ten people on the senior portfolio team | The April workshop was open to all staff and drew more than twenty attendees |
| Portfolio oversight | Portfolio companies were using or trying AI in different ways, as Maven described it | Portfolio directors had questions they could raise with portfolio companies, supplied in the follow up PDF |
| Client response to the first session | One talk delivered at a portfolio away day | Maven invited Insightful AI back for a second workshop, and further conversations followed |
| Client response to the second session | No feedback recorded at that point | Maven's compliance team reported positive feedback from attendees, in one message after the April workshop |
What the client said
The Team were really interested in the insights that you shared and there is clearly a need for Maven and our portfolio companies to implement a compliance framework and set of policies to address AI utilisation.
I wanted to say thanks again for delivering the session to the team, it was very helpful and we've had a lot of positive feedback on this.
What we learned
Training lands best when it starts in plain English. Maven's team did not need claims about what AI might become. They needed a clear account of what it does, where it fails, and why the judgement stays with people.
Portfolio teams want questions, not principles. The second workshop turned AI risk into questions a director could ask in a board meeting. That made the session useful well beyond Maven's own walls.
One session is rarely the whole job. The first talk created enough interest for Maven to ask us back. The second reached a wider group and was followed by written guidance they could keep.
Who is this for?
This type of AI governance training suits organisations whose senior teams need to understand AI risk before staff use gets too informal or too scattered.
- private equity firms
- portfolio directors and investment teams
- regulated organisations
- leadership teams overseeing AI adoption across several business units