This week’s dominant theme is the rapid shift from AI experimentation to autonomous, agentic execution across finance — from systematic trading platforms democratising quant strategies to wealth managers deploying AI agents that actually execute transactions. Simultaneously, the UK and EU regulatory apparatus is tightening, with the ESRB issuing a first-of-its-kind warning on frontier AI cyber risks to the financial system, and insurers facing an uncomfortable reckoning over trillions in ‘silent’ AI liability sitting unpriced inside conventional policies.
Top story: Insurers have more than 90% of their AI exposure in ‘silent’, largely unpriced cover buried inside conventional policies — a systemic liability gap co-authored with Anthropic and OpenAI researchers.
Insurers Sitting on Trillions in Unpriced ‘Silent’ AI Liability
PYMNTS · Risk
A report from the Artificial Intelligence Underwriting Company, co-authored with researchers from Anthropic and OpenAI, found that more than 90% of insurers’ exposure to AI risk is embedded in ‘silent’ cover within conventional policies — meaning it is largely unpriced and in many cases unnoticed. Researchers warn that rising litigation signals a shift from chatbots generating bad information to AI agents taking consequential action, potentially exposing businesses to professional negligence and wrongful death claims. For underwriters and risk teams, the implication is stark: existing policy language was not written with autonomous AI in mind, and the repricing challenge is enormous.
https://www.pymnts.com/insurance/2026/insurance-industry-may-be-unprepared-for-agentic-ai-risks/
ESRB Issues First Systemic Warning on Frontier AI Cyber Threat to EU Finance
Slaughter and May / Financial Regulation Weekly Bulletin · Regulation
The European Systemic Risk Board issued a formal warning (ESRB/2026/3) on systemic cyber risks to the EU financial system stemming from frontier AI models — advanced general-purpose systems now capable of discovering vulnerabilities and generating working exploits at scale. The warning lands as the Bank of England’s 2026 H1 Systemic Risk Survey recorded 82% of surveyed banks, insurers and asset managers citing cyber-attacks as a top-five risk. For compliance and risk officers across European financial institutions, this is the clearest signal yet that frontier AI cyber risk has moved from theoretical to a formal macro-prudential concern requiring board-level attention.
Horizon Trade Launches Agentic Platform to Democratise Systematic Investing
FinTech Global · Tools
Horizon Trade, an agentic fintech backed by Entrée Capital, officially launched this week with a waitlist exceeding 23,000 traders — its pitch being to convert natural-language trading ideas into fully automated strategies without requiring quant researchers, bespoke code, or institutional data subscriptions. The platform targets the long-standing access gap in systematic investing, a discipline historically confined to well-capitalised hedge funds and prop desks. If it delivers, it represents a structural shift in who can run quant-style strategies — with significant implications for asset managers competing for retail and semi-professional capital.
https://fintech.global/2026/07/17/horizon-trade-launches-ai-platform-for-systematic-trading/
P&C Insurers’ AI Spend Fails Underwriters: 64-Point Perception Gap Revealed
FinTech Global · Strategy
New research from Federato based on 750 insurance professionals found a staggering 64 percentage-point gap between executive and frontline views on AI effectiveness: while 91% of executives believe they have real-time portfolio visibility, only 27% of underwriters agree. The report argues the problem is integration, not investment — fragmented AI deployments are generating executive confidence without operational impact. For PE-backed insurers and MGAs investing heavily in AI transformation, this data is a warning that ROI claims need to be pressure-tested at the underwriting desk, not just the boardroom.
https://fintech.global/2026/07/21/why-fragmented-ai-is-undermining-pc-underwriting/
UK Aviva Reports £233m AI-Assisted Fraud as Broker Threat Perception Doubles
Life Insurance International · Risk
Aviva reported more than 18,000 suspect claims worth £233 million across its brands in 2025, with AI cited as a key contributing factor — including faked car crashes, deepfake documents and exaggerated claims. A GlobalData survey of UK commercial insurance brokers found that AI now ranks as the fourth-greatest threat facing the sector in 2026, with 9.6% selecting it as the single biggest threat, nearly double the 5.2% who ranked it eighth in 2025. The rapid escalation of AI-enabled fraud in the UK market is forcing insurers to rethink claims verification workflows and invest in counter-AI detection at speed.