This semi-monthly column highlights news, government documents, NGO/IGO papers, conferences, industry white papers and reports, academic papers and speeches, and central bank actions on the subject of AI’s fast paced impact on the banking and finance sectors.
NEWS:
OCC grants preliminary approval to Erebor Bank, a Peter Thiel-backed startup focusing on crypto and AI, The Block, October 15, 2025. A national bank regulator has granted “preliminary conditional approval” for the venture capitalist Peter Thiel-backed Erebor Bank, which plans to serve the cryptocurrency and artificial intelligence sectors. The Office of the Comptroller of the Currency (OCC) granted approval on Wednesday. Comptroller of the Currency Jonathan Gould called Erebor the “first de novo bank to receive a preliminary conditional approval” since he started his role at the OCC in July. “Today’s decision is also proof that the OCC under my leadership does not impose blanket barriers to banks that want to engage in digital asset activities,” Gould said in a statement. “Permissible digital asset activities, like any other legally permissible banking activity, have a place in the federal banking system if conducted in a safe and sound manner.” Erebor is aiming to fill the gap of Silicon Valley Bank, a bank popular with startups and venture capitalists that collapsed in 2023. Erebor was founded in 2025 by Silicon Valley mainstays Palmer Luckey and Joe Lonsdale with backing from Thiel’s firm Founders Fund and Haun Ventures, according to reporting from the Financial Times. Like other projects associated with Thiel, Erebor is named after a mountain in author J.R.R. Tolkien’s book series “The Lord of the Rings.” In its application, Erebor said it would be a national bank that would provide both traditional banking and crypto-related products and services. According to the Financial Times, citing an unnamed source close to Erebor, the bank’s application did not receive “special treatment” from the Trump administration, despite Luckey, Lonsdale, and Thiel’s longstanding ties to the president.
What Investing in the Age of AI Will Look Like – Wall Street Journal via MSN, October 9, 2025. Technology keeps remaking the way Americans invest, from the introduction of the electric telegraph in 1844 to the advent of artificial intelligence in just the past handful of years. This last innovation seems to hold immeasurable potential to upend norms for information, markets and the economy. We wanted to know what WSJ readers think about the revolution unfolding. So we asked, How do you think AI will change investing in the future? We received hundreds of responses. Here are some of them.
Are We Underestimating How AI Will Change Private Markets? Colossus, August 25, 2025. …Bloomberg was a key catalyst for spiraling changes in the bond and equity markets. We think AI will drive a similar pattern of change in private equity and venture capital today. Firms working against the backdrop of more crowded and competitive markets will readily adopt AI for operational efficiency. As this technology becomes ubiquitous, it will create opportunities to invest in completely new ways. In a world where every private markets investor relies on AI, we believe that the enduring moats of relationships, data, and operational excellence will matter even more. Cheap data aggregation means only truly proprietary data will provide an edge. Democratized information means firms need unparalleled access and genuine trust to win the best deals. Maximizing operating leverage requires systems that can adjust to conditions in real time. We explore these themes through conversations with the best founders building in this space today.
WHITE PAPERS AND REPORTS:
Product Segment Report: As Artificial Intelligence Expands, The Need for Strong Partnerships With Data Vendors And Robust Data Process Capabilities Will Continue to Grow. October 2025. Burton-Taylor, the world’s leading financial market and information research and consulting firm, delivers an analysis of the Product Segment Report. The analysis is sufficiently detailed to allow the reader to gain an insight into the changes in the financial market data sector, how technology will continue to play a critical role in the demand for market data, and what the successful financial data vendors look like.
TabbFORUM Report: The State of AI in the Capital Markets 2025 Beyond the Hype, from ‘Cool to Core’. Artificial intelligence is reshaping capital markets, but unevenly. Firms have pushed past proofs of concept into live deployments across the front, middle, and back office. But data silos, governance hurdles, and hallucination risks remain persistent obstacles. This TabbFORUM report explores how AI is moving from cool novelty to core infrastructure, what’s working in production today, and where the next wave of adoption will determine lasting competitive advantage. A year ago, the first edition of this report highlighted a capital markets industry on the brink of transformation. Artificial intelligence had begun to move beyond proofs of concept, edging into production workflows across the front, middle, and back office. But adoption was uneven, and practical challenges. Data fragmentation and hallucinations remained serious roadblocks. Today, that transformation is no longer theoretical. Firms across the ecosystem are piloting AI tools and scaling them. In 2025, the conversation has shifted from “if” to “how fast.” Buy-side leaders are embedding AI across investment workflows, broker-dealers are accelerating surveillance automation, and infrastructure providers are redesigning architectures to support agentic orchestration and retrieval-augmented generation (RAG). Even the regulatory landscape is evolving to meet these capabilities head-on. While hype persists, practitioners are increasingly skeptical of general-purpose platforms. They want traceability, explainability, and ROI. And they are rapidly developing internal governance models to match. What emerges from this report is a market no longer wondering whether AI can work. It already does. The question now is: Who will get it right, at scale, and with lasting strategic advantage? This report explores how advancements in AI is revolutionizing every aspect of capital markets, from trading strategies to compliance and portfolio management including [Download the Report]:
- Buy-Side vs. Sell-Side Approaches to AI
- AI Use Cases by Office Function
- Human-in-the-Loop vs. Full Automation
- Data Architecture & Infrastructure
- AI Governance, Regulatory Trends & Risk
- Agentic AI and Autonomous Systems
- Market Impact and Alpha Opportunities
Nasdaq launches new AI-ready data infrastructure aimed at the buy-side. The Trade, October 9, 2025. “Institutional teams are sitting on a wealth of data, but too often it’s locked behind manual processes and fragmented systems […] It’s about giving distribution teams the intelligence they need, when and where they need it,” Nasdaq’s Daniel Brickhouse tells The TRADE. Nasdaq eVestment has launched a set of AI-ready data infrastructure, to enhance institutional intelligence and allow institutional asset managers to activate agentic workflows directly within their own environments. The offering, which includes Nasdaq eVestment’s AI-ready datasets and its Next Best Action for Institutional Capital solution, is expected to remove the need for manual data manipulation and address fragmentation, to allow firms to make more informed market-speed decisions. Speaking to The TRADE, Daniel Brickhouse, vice president and head of product at Nasdaq Analytics, said: “Institutional teams are sitting on a wealth of data, but too often it’s locked behind manual processes and fragmented systems. “With AI-ready datasets and embedded decision support, we’re helping asset managers surface the right opportunities faster, whether that’s identifying under-allocated investors, spotting mandate risk, or prioritising outreach. It’s about giving distribution teams the intelligence they need, when and where they need it.” The launch also aligns with increasing pressure faced by firms and asset managers to respond to institutional mandates with speed and precision, as well as an increased interest in automation and AI integration into workflows across the industry, with nine in ten investment advisors planning to implement AI workflows in 2026, according to Nasdaq. Specifically, the optimised datasets span 27,500 strategies and 25,500 investor profiles, which can be licensed and integrated via delivery channels such as Snowflake and secure APIs. In September, Nasdaq expanded its strategic technology partnership with Amazon Web Services to provide the capability for financial institutions to deploy its Calypso platform onto AWS. Moreover, in the same month, the exchange submitted a filing to the US Securities and Exchange Commission (SEC), in a move which if introduced, will enable Nasdaq member firms to trade tokenised versions of equity securities and exchange traded products (ETPs) as regular securities. “Institutional teams are sitting on a wealth of data, but too often it’s locked behind manual processes and fragmented systems […] It’s about giving distribution teams the intelligence they need, when and where they need it,” Nasdaq’s Daniel Brickhouse tells The TRADE.
Via betanews: A new report from The Conference Board and ESGAUGE [The full report AI Risk Disclosures in the S&P 500: Reputation, Cybersecurity, and Regulation 03 October 2025. This report analyzes how the largest US public companies disclose artificial intelligence (AI) risks in their 2023–2025 annual filings, providing insight into the issues shaping board agendas, investor expectations, and regulatory oversight in the years ahead. The report finds that 72 percent of S&P 500 companies now flag AI as a material risk in their public disclosures. That’s up from just 12 percent in 2023, underscoring how rapidly AI has moved from experimental pilots to business-critical system. Reputational risk tops the list, cited by 38 percent of companies. Firms warn that failed AI projects, missteps in consumer-facing tools, or breakdowns in service could quickly erode brand trust. Cybersecurity risks follow, disclosed by 20 percent of firms. Unlike reputational or cybersecurity risks, which can manifest quickly, legal risk is framed as a longer-tail governance challenge that can lead to protracted litigation, regulatory penalties, and reputational harm.
We’re seeing a clear theme emerging across disclosures: Companies are worried about AI’s impact on reputation, security, and compliance. The task for business leaders is to integrate AI into governance with the same rigor as finance and operations, while communicating clearly to maintain stakeholder confidence,” says Andrew Jones, author of the report and principal researcher at The Conference Board. Finance, healthcare and industrial sectors have seen the biggest rise in disclosures. From 2023 to 2025, the number of companies disclosing AI-related risks jumped in financials (from 14 to 63 companies), healthcare (from five to 47), and industrials (from eight to 48). Why these sectors? Financial and health care companies face regulatory and reputational risks tied to sensitive data and fairness, while industrials are scaling automation and robotics. Intellectual property, privacy, and adoption risks are now surfacing too. 24 companies highlight risks spanning copyright disputes, trade-secret theft, and contested use of third-party data for model training. 13 companies warn of sensitive exposure under the General Data Protection Regulation, Health Insurance Portability and Accountability Act, and California privacy laws (CCPA/CPRA). Technology adoption is cited by eight companies, pointing to risks in execution such as high costs of new platforms, uncertain scalability, and the possibility of under-delivering on promised returns…”
See also Gallup – “Most Americans Expect AI Attacks From Foreign Governments. As AI continues to develop, many questions remain over its future capabilities. Even so, 87% of Americans say it is at least somewhat likely that foreign governments will use AI to attack the U.S. within the next two decades, including 43% who say this is very likely…”
PAPERS – BIS
Exploring household adoption and usage of generative AI: new evidence from Italy. Leonardo Gambacorta, Tullio Jappelli and Tommaso Oliviero. BIS Working Papers. 10 October 2025. Generative artificial intelligence (AI) has spread rapidly since late 2022, with tools such as ChatGPT and Google Gemini entering everyday life. People use them for shopping, study, entertainment or even financial and medical advice. While firms’ adoption of AI has been widely studied, much less is known about how households engage with these new technologies. Understanding household awareness and use is important because it shapes how AI affects skills, income and inequality. We contribute to the literature by presenting results from a comprehensive nationally representative Italian survey of gen AI adoption. We use a new module of the Italian Survey of Consumer Expectations, which was fielded in 2024. The survey covers 5,000 individuals and collects information on AI knowledge, actual and expected use, and socioeconomic characteristics. We not only identify which groups of people are more likely to use AI and for what purposes but also provide new evidence on the economic return of adoption by linking AI use to labour income. Findings – By mid-2024, three out of four Italians aged 18–75 had heard of generative AI, and about one in three had tried it in the past year. Around one in five used it every month. Awareness and use are higher among men, younger people, students and those with more education. Income matters less, and living in a big city does not make much difference. Looking ahead, people expect to use AI more for leisure and education than for financial or medical advice. We also find that using AI is associated with higher income. On average, users earn about 2% more than non-users. This is similar to the return from half a year of additional schooling. However, men appear to benefit more, suggesting that AI may widen existing gender gaps in the labour market. We present findings from a specialized module on generative artificial intelligence (gen AI) included in the Italian Survey of Consumer Expectations (ISCE), conducted in 2024 with a representative sample of Italian individuals. This analysis offers novel insights into current and anticipated interactions with gen AI tools and the potential benefits from adoption. As of April 2024, 75.6% of the Italian population aged 18–75 was aware of gen AI, 36.7% had used it in the previous 12 months, and 20.1% reported monthly usage. Socio-economic factors significantly influence adoption rates, with higher usage observed among men, individuals with college degrees, and younger individuals, particularly students. Looking ahead, gen AI is expected to be used more frequently for education and leisure activities in the coming months. Finally, using a Mincer earnings regression, we highlight that the income return associated with gen AI usage is around 2%.
PAPERS – NBER
Artificial Intelligence in Research and Development. Benjamin Jones. Working Paper 34312. DOI 10.3386/w34312. October 2025. How much can AI accelerate progress in different research fields? This paper shows that three features—the share of research tasks AI performs, the productivity of AI at those tasks, and the strength of bottlenecks—are key determinants of AI’s implications in any area, from cancer therapeutics to software design. The model maps changes in AI capabilities to research outcomes, quantifies the “marginal returns to intelligence,” and shows how AI can shift returns to R&D investment. Concepts like superintelligence, Powerful AI, and Transformative AI are further engaged and disciplined. Finally, the framework sets a measurement agenda linking AI benchmarks to field-specific opportunities for accelerating progress.
The Impact of AI and Digital Platforms on the Information Ecosystem.Working Paper 34318. DOI 10.3386/w34318. We develop a tractable model to study how AI and digital platforms impact the information ecosystem. News producers — who create truthful or untruthful content that becomes a public good or bad — earn revenue from consumer visits. Consumers search for information and differ in their ability to distinguish truthful from untruthful information. AI and digital platforms influence the ecosystem by: improving the efficiency of processing and transmission of information, endangering the producer business model, changing the relative cost of producing misinformation and altering the ability of consumers to screen quality. We find that in the absence of adequate regulation (accountability, content moderation, and intellectual property protection) the quality of the information ecosystem may decline, both because the equilibrium quantity of truthful information declines and the share of misinformation increases; and polarization may intensify. While some of these problems are already evident with digital platforms, AI may have different, and overall more adverse, impacts.