This is a Reach announcement and the information contained is not considered to be material or to have a significant impact on management's expectations of the Company's performance.

20 April 2026

Insig AI plc

("Insig AI" or the "Company")

Insig AI demonstrates Central Bank MCP server in action: first published study reveals how 55 central banks talk about war

Further to its announcement of 13 April 2026, Insig AI plc (AIM: INSG), a leading provider of AI-led analytics and machine-learning solutions, is pleased to provide an update regarding the launch of its Central Bank Model Context Protocol ("MCP") server. Following initial client interest, the Company has published the first in a series of demonstration studies built entirely through the MCP server, titled Conflict Language: How Central Banks Talk About War. The interactive analysis is available to view at: insights.insig.ai/i/central_bank_geopolitical_language_apr15 and evidences, in a single worked example, the speed, depth and scale of insight the MCP server makes possible.

A Live Demonstration of the MCP Server

The study was produced end-to-end through a natural-language conversation with Anthropic's Claude, using the Insig AI Central Bank MCP server as the data layer. Claude queried the dataset, identified the keywords and phrases most strongly associated with wartime policy communications, grouped them by linguistic category, and rendered the results as an interactive, publishable dashboard - in a single session.

Work of this scope would traditionally require a research team several weeks of manual document review, Structured Query Language questioning and dashboard engineering. With the MCP server, the full pipeline - from question to published insight - was completed in a fraction of the time, without intermediaries, bespoke integrations or custom Business Intelligence tooling.

Addressing Increasing Demand for Real-Time Policy Insight

Demand from institutional investors for rapid, structured analysis of central bank communications has grown sharply as markets navigate overlapping conflicts, divergent policy paths and tariff-driven supply shocks. Tone and terminology - not only headline rate decisions - are now actively priced by macro, fixed income and risk teams.

The Conflict Language study is the first of a planned series of demonstration analyses designed to showcase the range of questions the MCP server can answer: from monetary policy divergence, to forward guidance tone, to cross-bank reactions and to specific events. It is also enhancing the Company's ongoing commercial engagement with asset managers, hedge funds and research institutions.

What the Analysis Shows

The study tracks 31,321 mentions of 76 war- and crisis-related keywords across 4,235 statements, minutes and speeches from 55 central banks between January 2022 and April 2026. Key findings include:

•        "Financial stability" is the most-used term across the dataset, with 6,069 mentions, followed by "volatility" (4,058), "energy prices" (3,975) and "commodity prices" (2,932).

•         The April 2026 data refresh recorded 154 new mentions across 16 documents from 15 banks in a single update, led by Price Pressure (+92) and Financial Stress (+30) language - driven by the US/Israel/Iran conflict and its pass-through to energy and commodity markets.

•           Keyword spikes map directly onto named geopolitical events: sanctions, supply chain and wheat surge around the Ukraine invasion; emergency and second-round effects spike around Gaza; safe haven and heightened uncertainty climb through tariff escalations.

•           The European Central Bank (4,858 mentions), Bank of England (2,739) and Federal Reserve (2,330) lead by volume, while emerging-market central banks display distinct category signatures - particularly around commodity and supply-side shocks.

Richard Bernstein, Chief Executive of Insig AI, commented:

"Last week's MCP announcement set out its capability. This study is an example of its outcomes. In a matter of hours, our team used the Central Bank MCP server together with Claude to interrogate over 4,000 documents across 55 central banks and produce a publishable piece of research that would previously have taken a sell-side research desk several weeks to deliver. That is the shift in productivity we are now putting directly into our clients' hands. The early feedback from asset managers and macro investors has been extremely positive. We believe that the scale and speed of this offering unlocks a market edge."

For further information, please visit www.insig.ai or contact: