This week’s stories reveal AI crossing from experimentation into regulated execution across finance: the FCA’s landmark Mills Review found consumers already using AI for pension and debt advice, while Goldman Sachs backed a platform purpose-built to automate banks’ highest-stakes decisions. The underlying pattern is clear — agentic AI is no longer a future threat or opportunity, it is a present operating reality that regulators and capital allocators are now racing to govern and fund simultaneously.
Top story: The FCA’s Mills Review found one-in-five UK adults ready to use agentic AI for financial decisions — and called for urgent regulatory action within months.
FCA Mills Review: One-in-Five UK Adults Ready for Agentic Finance
Regulation Tomorrow / FCA · Regulation
Published on 6 July 2026, the FCA’s landmark Mills Review — the first of its kind globally — found that 29% of UK consumers who sought pension advice in the past year used AI to do so, and that one in five UK adults (roughly 11 million people) are open to agentic AI making financial decisions on their behalf. The review made seven priority recommendations, including a perimeter review — to be completed within three to six months — examining whether general-purpose LLMs providing ‘advice-like support’ on savings, pensions and mortgages fall outside the regulatory boundary and require new rules. For practitioners, the key message is accountability: the Senior Managers Regime still applies even where AI models and their updates sit partly outside a firm’s direct control.
Goldman Sachs Leads $110M Into AI Decision Platform for Banks and Insurers
Fortune · Finance
Goldman Sachs Growth Equity led a $110M Series C into Taktile, an agentic decision platform that automates high-stakes regulated decisions — including loan approvals, AML triage, insurance claims, and customer onboarding — for banks and insurers. One of the world’s largest insurers is already running multiple use cases on Taktile with projected cost savings exceeding $90M in claims processing alone, and clients report 95% automation in B2B underwriting and a 75% reduction in AML false positives. The bet signals Goldman’s view that frontier AI has crossed a capability threshold where it can reliably handle decisions that previously required human experts — and that the next competitive divide in financial services is who can operationalise agents at scale within regulated environments.
https://fortune.com/2026/06/24/exclusive-taktile-goldman-sachs-ai-bank-insurance-funding/
CB Insights: Agentic Trust Infrastructure Funding Already Exceeds Full-Year 2025
Fintech News Singapore · Generative AI
A new CB Insights market map of the AI fraud defence sector found that year-to-date equity funding for agentic trust infrastructure companies in 2026 already exceeds full-year 2025 totals, driven by surging AI-initiated fraud and the need to govern autonomous agents operating inside financial systems. Notable deals include Oasis Security’s $120M Series B for AI agent access management — serving clients including Citizens Bank — alongside a wave of platforms focused on real-time fraud detection, digital identity verification, and agent-level policy enforcement. For financial services practitioners, the finding reframes AI fraud defence not as a compliance line-item but as a fast-growing infrastructure category attracting institutional venture capital at pace with the agentic AI deployments it is designed to secure.
https://fintechnews.sg/133611/ai/top-companies-in-the-ai-fraud-defense-market-in-2026/