AI in Finance Summit London 2026

The AI in Finance Summit London is RE·WORK’s City-focused event on applied machine learning in financial services. The official page gives the date as 15 September 2026, at 1 America Square in the City. The venue (Cavendish Venues, 17 Crosswall, EC3N 2LB, per the venue’s own site) is a few minutes’ walk from Tower Hill, Aldgate and Fenchurch Street, and has a stretch of the Roman London Wall running through it — a more memorable backdrop than the usual conference-centre basement. It is co-located with CDAO Financial Services, so a single trip covers both the applied-AI and the chief-data-officer conversations.

RE·WORK’s house format is worth understanding before you buy a ticket. These are not sprawling expos; they are tightly curated, presentation-led days with a strong bias towards practitioners describing systems they have actually shipped, followed by questions and structured networking. The 2026 edition runs multiple content streams, group problem-solving sessions with expert attendees, and question-and-answer slots with speakers, plus a closing reception; a PLUS Pass upgrade gives access to post-event recordings of the sessions, which is the pragmatic option if you want the content but not the day out of the office.

The topic list is squarely operational rather than speculative: financial forecasting, fraud prevention and detection, anti-money laundering, risk management, model risk, regulation, portfolio optimisation, wealth management, retail and wholesale banking, conversational AI, natural language processing and fintech. The through-line in the 2026 line-up is the shift from predictive models to agentic systems — and, crucially, to the governance layer around them.

Speakers listed on the official site at the time of writing include Ryan Courtier, Senior Vice President and AI Product Leader at Citi, and Ramy Erfan, Vice President for Business and Technology Enablement, also at Citi; Parag Kumar, Executive Director and Data Analytics Lead for Compliance, Conduct and Operational Risk at J.P. Morgan Chase; Rajasree Ramakrishnan, Product Lead for Agentic Observability at Lloyds Banking Group; and Caroline Cartellieri, a Non-Executive Board Member at Saffron Building Society. The wider published list adds Charles Phiri, Executive Director for SME AI/ML Innovation at J.P. Morgan Chase; Annapoorani Lakshmi Narayanan, Product Manager Lead for AI Research at the same bank; Soung Low, Model Risk Data Scientist at NatWest Group; Kaushik Chaubal, Senior Director for Investor Solutions at BlackRock; Amit Kumar, Associate Director and eFx Quant at HSBC; Parinita Kothari, Engineering Lead, and Eshaan Salwan, Cybersecurity Audit Analyst, both at Lloyds Banking Group; and Hemant Patkar, Cybersecurity Lead Designer and Architect at Virgin Money. RE·WORK states that more speakers are still to be announced.

Two things stand out in that roster. First, the job titles are unusually specific — “Product Lead, Agentic Observability” and “Model Risk Data Scientist” are roles that did not meaningfully exist in UK banking a few years ago, and their presence is a better indicator of where large institutions are actually investing than any keynote abstract. Second, the mix skews towards the control functions: compliance, conduct and operational risk, model risk, cybersecurity audit, observability. This is a summit about putting AI into regulated production and keeping it there, not about demos. IBM is the platinum sponsor.

The intended audience, per the organisers, is chief technology, data and financial officers, directors of applied AI, heads of data and analytics, principal data scientists, machine learning engineers, financial analysts and strategy leads — drawn from banking, insurance and the broader financial-services sector. If you are a practitioner inside a bank or insurer trying to work out what “agentic” means in a setting with model-risk governance and a regulator, this is a well-chosen day: the room is full of people wrestling with the same constraint, and the co-located CDAO track means you can also collar the person who owns the data estate. If you want research-grade machine learning, look elsewhere — this is an applied, industry event, and it is honest about that.