London, UK – 3 June 2026 – Guildhawk today announces the launch of its new GAI Labs website, now live at GAI Labs official site, marking a significant milestone in the evolution of its research and development arm dedicated to building trustworthy, regulated AI systems.
Gai Labs brings together Guildhawk’s in‑house team of coders, AI engineers, data scientists and linguistics experts, alongside specialists in multilingual workflows, health and safety, fraud prevention and regulatory compliance.
Gai Labs to unite expertise across business, academia, government, entrepreneurship, cyber security and OSINT to develop AI solutions that address real‑world safety risks.
The newly launched platform articulates Gai Labs’ founding thesis: that language is a system of trust, where meaning, authority and risk are inseparable - and where current AI systems often produce fluent outputs faster than they can be trusted.
Building AI that can be trusted in regulated environments
Gai Labs focuses on developing AI systems with traceability, auditability and human oversight built in from the start, supporting compliance with emerging global frameworks such as the EU AI Act.
Unlike general-purpose AI models prone to hallucinations and unverifiable outputs, GAI Labs takes a disciplined approach grounded in:
- Human‑verified data
- Domain‑specific models (Small Language Models)
- Audit trails and compliance controls
- Expert-in-the-loop validation
Its research and testing programmes are built around real-world failure modes observed in regulated industries, with the focus on guaranteeing AI outputs are reliable in sectors where errors can have legal, clinical or financial consequences.
Backed by academic research and Innovate UK
Gai Labs builds on Guildhawk’s long-standing collaboration with Sheffield Hallam University through Innovate UK–backed Knowledge Transfer Partnerships (KTPs), recognised among the top 50 in UK history.
These partnerships have delivered:
- The development of Gai Translate, with the world’s first secure 1-click human verification
- New approaches to multilingual dataset labelling
- Advancements in trustworthy AI built on high‑quality domain specific data verified by experts
Professor Alex Shenfield, Professor of Machine Learning at Sheffield Hallam University, said:
“High-quality data is the foundation of trustworthy AI. Our work with Guildhawk is helping to solve a critical bottleneck in the development of intelligent systems.”
Award-winning secure AI solutions
The Gai Labs team has already produced a suite of secure AI technologies, including Gai Translate, which won the Market Disruptor of the Year Award at the 2025 Security & Safety Entrepreneur Awards (SSEA).
Developed through extensive R&D in Sheffield, UK and backed by Innovate UK, the Gai Suite of products combines AI automation with expert human review, delivering secure and accurate multilingual communication for regulated industries.
A response to rising AI risk
The launch comes at a time when malicious actors are increasingly using AI, such as deepfakes and automated deception, to exploit individuals and organisations.
David Clarke, Chief Security Officer at Guildhawk and Founder of Gai Labs, said:
“We are at a tipping point. Bad actors are already using AI to deceive, defraud and harm. At Gai Labs, we are using AI to fight back. We are building systems that protect people, expose risk, and make the world safer.”
He added:
“Trust must be engineered into AI from day one. Without traceability, auditability and human oversight, AI creates risk. With them, it becomes a powerful tool for public good.”
Human oversight at the core of AI
Jurga Žilinskienė MBE, Founder and CEO of Guildhawk, emphasised the importance of retaining human accountability:
“We have no method of relying on AI output alone. To be able to rely on it, human decision must remain central.”
She added:
“Our mission is clear: to make the world safer with AI. We are building systems where humans remain accountable, where every decision can be verified, and where technology enhances, not replaces, expert judgement.”
A new category: language intelligence
The Gai Labs website highlights how Guildhawk is leveraging its 25 years experience in linguistics and multilingual data to what it defines as “language intelligence”, a discipline focused on managing risk, meaning and compliance in both monolingual and multilingual communication.
Its flagship research, includes the development of new agentic tool to prevent fraud that has the potential to open a new category of protective controls, showing how AI is set to become a game-changer for the world.
Media enquiries:
info@guildhawk.com
