Skip to content
Briefings are running a touch slower this week while we rebuild the foundations.See roadmap
Advai
OrganisationGB

Advai

London AI safety testing firm; technical partner in FCA AI Live Testing programme.

Last refreshed: 22 April 2026 · Appears in 1 active topic

Key Question

Does a regulator endorsement turn an AI-assurance startup into a standard-setter?

Timeline for Advai

#220 Apr

Appointed as technical partner for FCA AI Live Testing second cohort

UK Startups and Innovation: FCA names eight firms in AI Live Testing round two
View full timeline →
Common Questions
What does Advai do?
Advai is a London AI-assurance company that red-teams AI systems, probing them for failure modes, bias, and adversarial vulnerabilities. It serves financial services, critical infrastructure, and government clients.
Why did the FCA choose Advai as technical partner?
The FCA named Advai as technical partner for its second AI Live Testing cohort in April 2026 to provide independent adversarial evaluation of AI systems deployed by the eight cohort firms under live-market conditions.Source: FCA
What is AI red-teaming in financial services?
AI red-teaming probes models for edge-case failures, distributional shift, and deliberate adversarial manipulation before and during live deployment. In financial services, Advai-style assurance helps regulators and firms identify AI risks before they affect markets or consumers.
How is Advai involved in the FCA sandbox?
Advai serves as the technical partner for the FCA AI Live Testing second cohort (April 2026), stress-testing AI models deployed by Aereve, Coadjute, Palindrome, Barclays, Experian, GoCardless, Lloyds/Scottish Widows, and UBS.Source: FCA

Background

Advai was named as technical partner for the FCA's second AI Live Testing cohort on 21 April 2026, providing independent adversarial evaluation of AI systems deployed by Barclays, Experian, GoCardless, Lloyds/Scottish Widows, UBS, and four other financial firms . Advai's role is to stress-test AI models for failure modes, bias, and adversarial vulnerabilities before and during live deployment.

Advai is a London-based AI safety and assurance company. It operates a platform for red-teaming AI models: systematic adversarial testing that probes how AI systems respond to edge cases, distributional shift, and deliberate manipulation. The company works across financial services, critical infrastructure, and government sectors.

The FCA appointment elevates Advai from a specialist vendor to a regulator-endorsed AI assurance provider, a significant credibility signal in a sector where AI governance frameworks are nascent. As FCA Regulatory Sandbox applications rise 49% year on year, demand for Advai-style assurance is expected to scale in parallel.