This week’s stories reveal a finance sector caught between AI’s expanding attack surface and its untapped defensive potential: deepfake fraud is overwhelming insurers’ detection capabilities, regulators on both sides of the Atlantic are scrambling to future-proof their frameworks, and new AI underwriting partnerships signal that automation is moving from pilots to production pipelines. The common thread is that AI is now both the threat and the remedy — and the institutions that understand the difference will define the next era of financial services.
Top story: Aviva data exposes a deepfake fraud crisis: 98% of insurers hit by AI-altered claims evidence, yet only 32% confident in their detection capability.
Deepfake Claims Crisis: Insurers Overwhelmed by AI-Generated Fraud Evidence
Digital Insurance · Risk
A landmark Verisk State of Insurance Fraud Study found that 98% of insurers report AI editing tools are actively fuelling digital fraud, yet only 32% feel confident they can detect AI-generated deepfakes in claims submissions. Aviva’s own data showed it detected approximately £233 million in suspected claims fraud in 2025, with fraudsters increasingly using AI-generated images and manipulated documents to support false or exaggerated claims. For claims teams and fraud leads, the implication is stark: a single photo, document, or receipt can no longer be taken at face value, forcing a fundamental rethink of evidence verification infrastructure.
https://www.dig-in.com/opinion/why-ai-fraud-tools-could-cause-claim-risks
Sixfold and Sollers Partner to Automate AI Underwriting Across Europe and UK
FinTech Global · Tools
European insurtech Sixfold and insurance IT consultancy Sollers have announced a joint partnership to roll out AI-powered underwriting automation across Europe, the UK and North America. The collaboration enables carriers to automate submission intake, evaluate risk against their own appetite, and receive clear, explainable underwriting recommendations — targeting the slow, manual triage that has long bottlenecked commercial lines. For practitioners, this signals that AI underwriting is crossing from proof-of-concept into production-scale deployment, with explainability built in from the start to meet regulatory expectations.
https://fintech.global/2026/08/14/sixfold-and-sollers-team-up-on-ai-underwriting/
UK FCA’s Mills Review Flags AI as Systemic Driver Reshaping Finance to 2030
Global Regulation Tomorrow · Regulation
The UK’s Financial Conduct Authority has published the Mills Review, identifying AI as a systemic driver of change across financial services through 2030 and describing four distinct ‘system shifts’ already underway. The review sets out seven priority recommendations, including securing and adapting the regulatory perimeter as autonomous AI systems blur the boundaries of regulated activity. For UK-based banks, asset managers and insurers, the review is a significant signal that the FCA’s principles-based approach is evolving — and that firms should expect sharper supervisory attention on AI governance even without new prescriptive rules.
AI Trading Bots Go Mainstream — But Can Retail Investors Really Beat Hedge Funds?
Bloomberg · Finance
Bloomberg reports that a wave of AI-powered trading tools is enabling retail investors to build and deploy automated options-trading strategies previously exclusive to well-funded quantitative hedge funds. While enthusiasm is high, early evidence suggests retail AI bots are struggling to consistently outperform simple index fund strategies — raising questions about whether democratised algorithmic trading creates real alpha or merely new risks. For institutional players, the story matters because it signals both a shifting competitive landscape and the potential for increased market noise as amateur AI-driven flows scale.
‘Algorithmic Herding’: Bank of England and IOSCO Name AI Herd Risk as Live Systemic Threat
One Day Advisor / IOSCO · Risk
The Bank of England and IOSCO have formally classified ‘algorithmic herding’ — where AI systems at different firms converge on correlated trading decisions — as a live financial-stability concern, not a theoretical future risk. The warning emerges as regulators observe heavy AI adoption in institutional research and trading, while noting that real caution persists about letting AI make unsupervised decisions. For risk officers and CROs at banks, asset managers and insurers, the message is clear: the systemic risk posed by multiple firms running similar AI models simultaneously is now on regulators’ watchlists and firms should expect questions about model diversity and concentration.
https://www.onedayadvisor.com/2025/07/ai-in-algorithmic-trading-investment.html