The conversation around AI in financial services has moved from "what if" to "what happened when we actually tried it"—and that shift was on full display at FStech's Future of AI in Financial Services conference on 9 October 2025.

Held at a packed London Hilton Tower Bridge, the second edition of the event brought together senior leaders from banks, insurers, FinTechs and regulators who came bearing battle scars, production systems and a refreshing willingness to discuss what didn't work alongside what did.

Opening the day, Jonathan Easton, editor at FStech, set the tone. "Last year we explored the promise; today we're here for the practice," he said. "We're moving from 'maybe we could' to 'here's what we did' – and more importantly, here's what actually worked."

Gone were the glossy vendor pitches. In their place: Jeff Valane, group head of AI management and strategy at HSBC, discussing the realities of building an AI native bank, including the joy of explaining AI model risk to boards and discovering that "just deploy it" is not a viable governance framework.

The morning also featured a glimpse at serious computational firepower. Dr Richard Gilham, AI supercomputing infrastructure specialist at the Bristol Centre for Supercomputing, walked delegates through Isambard-AI, the UK's most powerful supercomputer, and what high-performance computing actually means for a sector still figuring out where cloud ends and cutting-edge research begins.

Agentic AI—autonomous systems that can make decisions and execute multi-step processes—emerged as a key theme. A panel featuring Ramona Fuchs, chief executive and co-founder at Moterra, alongside leaders from AXA UK & Ireland and First Direct Bank, tackled the thorny questions around regulatory compliance, accountability and legacy system modernisation.

But perhaps the most pressing issue was people, not machines. A panel sponsored by Smarsh explored how organisations are reskilling workforces for an AI-augmented future, with contributions from Allianz Global Investors, Standard Chartered and Innovate Finance.

Professor Karen Elliott from the University of Birmingham grounded the discussion in corporate digital responsibility, arguing that quality assurance and human oversight remain vital as companies move from isolated AI agents to orchestrated multi-agent teams.

After lunch, attention turned to operational resilience. Dalvinder Kular, assistant editor at FStech, moderated a panel examining how institutions balance innovation with compliance, featuring discussions on AI-driven stress testing, cybersecurity and governance frameworks from speakers including representatives from Queen Mary University of London, Forvis Mazars and BNP Paribas.

The day closed with Oscar Barlow, head of AI advocacy at Starling Bank, sharing candid insights from Starling's partnership with Google Gemini and the launch of its Spending Intelligence tool—a rare look at embedding ethical governance whilst building an AI-first culture.

Reflecting on the event, Easton said: "What made today special was the willingness of leaders to share not just their successes but their stumbling blocks. We're past the point of AI being a distant possibility—it's here, it's being deployed, and the industry is learning fast what works in production versus what looks good in a pitch deck."

Delegates left with practical lessons, new connections and, for one lucky winner, an Apple Watch.

FStech’s next conference, Cybersecurity Live, takes place November 6th at the Hilton London Tower Bridge. Fox Ahmed, global head of cybersecurity and technology and data protection regulatory risk, BNP Paribas has already been announced as one of the keynote speakers, with many more still to be revealed. Register here.

---