The information contained within this announcement is deemed by the Company to constitute inside information stipulated under the Market Abuse Regulation (EU) No. 596/2014, as retained as part of the law of England and Wales. Upon the publication of this announcement via the Regulatory Information Service, this inside information is now considered to be in the public domain.

FOR IMMEDIATE RELEASE

01 October 2025

London, UK

Defence Holdings PLC

("Defence Holdings" or the "Company")

Defence Technologies Announces Second Classified AI Product Build: Edge Analysis & Identification

Defence Holdings PLC (LSE: ALRT), the UK's first listed software-led defence company, today announces Defence Technologies' second classified AI product build. Following the launch of Project Ixian in September, the second product is focused on edge-based analysis and identification support, one of the highest priority requirements for UK and allied forces. In simple terms, "the edge" refers to deployed environments, where data is captured and decisions are made in real time, often in austere or disconnected conditions.

Key Highlights

●    Second classified AI product build confirmed under Defence Technologies, focused on edge analysis, identification and decision-support capability, addressing one of NATO's most urgent priorities.

●    Addresses a critical operational gap: improving identification, reducing misclassification, and supporting decision making in contested environments.

●    Software-led solution extends the value of existing hardware fleets by processing data at the edge and synchronising outputs back to headquarters, maximising ROI on deployed hardware and reducing procurement costs.

●    Fully aligned with the UK Strategic Defence Review 2025 ("SDR25") and NATO interoperability objectives, giving Defence Holdings a direct policy tailwind and adoption pathway.

●    First classified build announced since the appointment of Field Marshal Lord Houghton as Incoming Non-Executive Chairman.

Product Overview

This capability is software-led and AI-enabled, designed to extend the value of existing hardware. It ingests data from deployed sensors and moves time-critical AI/ML inference onto secure edge devices. AI models fuse multi-sensor inputs, flag anomalies, and prioritise actions, cued to autonomous systems with a human in/on the loop, so operators and headquarters receive timely, decision-ready context. The result is improved identification confidence, faster cycles, and reduced operational risk, helping teams stay ahead of an adversary's OODA loop.

Practical examples include on-platform sensor fusion (EO/IR, radar and acoustic) to improve recognition and de-confliction in seconds; uncrewed systems performing onboard detection and route re-planning to reduce bandwidth and latency; military units correlating radar and electro-optical tracks at the edge to enhance identification in cluttered littorals; and forward elements using handheld edge devices to triage ISR, logistics and maintenance data before synchronising summaries to the enterprise.

For investors, this demonstrates the repeatability of Defence Holdings' sovereign AI model, rapidly converting strategy into a portfolio of products with direct commercial pathways across UK and allied markets.

The programme builds directly on Defence Holdings' strategy to assemble a portfolio of sovereign, AI-enabled products aligned with SDR25 priorities, from information operations to Intelligence, Surveillance, and Reconnaissance.

Defence Holdings will be working alongside a new hyperscale partner, underlining the Company's ability to pair sovereign AI capability with the reach and resilience of the world's most trusted technology platforms, while strengthening its position at the centre of an emerging global ecosystem for sovereign defence technology.

Leadership Commentary

Andrew McCartney, Chief Technology Officer of Defence Holdings, commented:

We're breaking through the traditional defence cycle, and the momentum we're building is already visible across the community. Ixian demonstrated that sovereign AI can be conceived, built, and aligned to live defence priorities in weeks and months, not years. This second classified build is proof that the model is scalable into areas of the highest operational value, such as identification and edge decision support. These are not niche requirements but critical problems for the UK and allied forces. For Defence Holdings, it represents both a significant commercial opportunity and the chance to establish ourselves as the sovereign partner of choice in some of the most strategically important areas of defence, a market expected to expand significantly as NATO and allies prioritise sovereign digital capability.