October Product Updates: UK PEP Data, Adverse Media & MCP

Every screening review depends on the information behind it. This month, we've added an enriched UK PEP dataset and expanded Adverse Media search.

Alongside those updates, we're taking a closer look at how Risk Inference helps your team interpret screening information and how Minerva MCP connects with the AI tools you use.

Key updates

1. New UK PEP dataset

We've added 3,600+ UK politically exposed person (PEP) profiles through our own enriched dataset. Available in the Minerva app and through the API, it adds a Minerva-built source to our existing PEP coverage.

The dataset covers current and former MPs, members of the House of Lords, devolved-government officials and other UK PEPs.

For teams screening customers with UK connections, this brings another dedicated source into their checks. Our PEP policy explains how we use public roles, jurisdiction and time in office to determine which records belong in PEP screening.

2. More precise Risk Inference

Risk Inference identifies political exposure, criminal risk and high-risk industries from screening information. Its Decision Classifier engine uses an AI model to assess information Minerva has already gathered: the person or company's role, organization, industry and sourced notes. It considers that context together to select from the risk labels configured in your workspace.

That includes public-web research from the Open Source feed. The classifier can recognize a public role described in different wording or another language, while reducing false positives caused by isolated words.

In the screening result, reviewers see Risk Inference, the selected risk category and the context supporting it. Your team can see why a finding was raised and review the underlying information as part of its decision.

An administrator can select Decision Classifier in Administration → Configuration → Risk Inference. The configuration guide explains the available settings and required search feeds.

3. Expanded Adverse Media search

A relevant article adds evidence about your customer. An article about somebody with a similar name adds work. This month's changes improve both news discovery and name matching.

We've added a search method that uses names tagged within news articles to find more reporting about the person being screened. It runs alongside the existing news search, expanding the articles returned for review.

More precise name matching reduces results about other people. Together, these changes bring your team more relevant coverage with fewer unrelated articles to sort through. These improvements are live in the News (Adverse Media) feed.

4. Connect your AI tools with Minerva MCP

Minerva MCP gives compatible AI tools a direct connection to Minerva. Your connected tool can run screening searches and read the results, or retrieve an existing customer profile, its potential matches and comments.

For a team using an AI tool during research, this brings Minerva information into that work without manually copying customer records into the conversation.

An administrator first enables MCP and chooses the permitted actions. Access follows your Minerva user permissions. The MCP setup guide walks your technical team through connecting a compatible tool.

To use the UK PEP dataset in the app or through the API, enable Decision Classifier or connect your AI tools through MCP, contact support@gominerva.com.