INTRO

I've added quite a few new subscribers lately, in part from some recent recommendations by James Falbe and Alexander Izydorczyk. (Thanks for the shout out guys!)

So I want to summarize some AI + finance themes I’ve been writing about and some other analysis I have in the works. I’m breaking it out this way in part because you can't easily track what I've covered on my website—an issue I plan to fix this fall with a new layout. Here we go:

The cost of using AI is going down dramatically

• It’s been well covered in the media that building and training AI is expensive, but the price to use it is dropping.

• In two years, it’s fallen 90% and shows no signs of slowing.

  • In a podcast that I host with Yale’s Francesco Fabozzi, Tharsis Souza, former Senior VP of Product Management at Two Sigma, told us earlier this year: “The price of intelligence is going to zero.”

The cost to extract data from documents is dropping as well

• LLMs can organize messy data, so you can take a bunch of PDFs and get text back. In jargon-y terms, you can structure unstructured data.

  • Right now, it’s hard to do this at scale with high precision with LLMs alone. Companies such as LlamaIndex pair traditional OCR with AI: once text is captured through OCR, LLMs can structure and analyze it. More to come on this.

AI Adoption is …. Slow

• This is a topic I write about like every week (including this one.) AI adoption is slow across financial services, though there’s growing consensus that you need to adopt.

Beware of look-ahead bias in academic research

• LLMs are black boxes, meaning researchers testing financial scenarios don’t know exactly what the model was trained on, so it may have seen the answers to the test already.

  • “These models aren't thinking through financial statements - they're remembering patterns from their training data,” Bradford Levy, University of Chicago professor, told me earlier this year.

• That said, I’ve covered some fun/interesting papers like how AI can uncover hidden drivers of stock returns or ranking the top AI models for stock, ETF, and crypto trading.

AI’s recent advances surprise even the experts

• I’ve chatted with dozens of experts over the last year and they're all pretty amazed by AI’s progress. No one really expected this.

• The unsettling part is that no one knows exactly how AI works.

  • Dario Amodei, a co-founder of the AI company Anthropic, wrote in April: “People outside the field are often surprised and alarmed to learn that we do not understand how our own AI creations work. They are right to be concerned: this lack of understanding is essentially unprecedented in the history of technology.”

  • I’m actually working on a feature for Chicago Booth Review on this topic, which is oddly called “interpretability.”

• This is not an academic exercise. AI going rogue is a real problem.

  • Replit's CEO apologizes after its AI agent wiped a company's code base in a test run and lied about it, Business Insider reported last week.

Interviews

Want to hear from more voices? I've interviewed executives from Moody's, Skadden, Snowflake, and other industry leaders, check them out here.

In Closing

There are, I’m sure, other topics that are worth mentioning but these are most top of mind. If I missed any or you’d like to suggest some topics, just reply to this email. It’s always nice to hear from readers.

FUNDRAISING

Perplexity Partner Quartr Raises $10M

Quartr, a Stockholm-based company that aggregates investor relations materials from public companies worldwide, has raised $10 million from existing investors Altos Ventures, Yanno Capital, and Öhman, to support expansion and product development.

The company operates two businesses: a data licensing API that feeds structured earnings calls and investor presentations to AI platforms and trading systems, and Quartr Pro, a direct research platform sold to asset managers, portfolio managers, and investor relations teams.

"The genesis of Quartr was really our founders were guys on Twitter who were talking investments and they always had this problem where they had to go to like a million different places to get IR material," Andres Sehr, a company spokesperson, told me. "So they thought, why isn't there like a Spotify for IR material—a one-stop shop where I can get all of that?"

The API business has secured partnerships with generative AI firms, most notably Perplexity AI, which integrates Quartr's structured earnings call data to enhance financial query responses. Other data customers include Rogo, and Bigdata.com. The company also powers live earnings call features on Yahoo Finance and several Nordic brokers.

Quartr Pro serves institutional clients directly, with customers including three of the world's five largest hedge funds by assets under management and two of the "Magnificent Seven" U.S. tech firms according to a press release. Some clients purchase both products—using the API for custom integrations into quantitative models while also accessing Quartr's user interface for research.

"We even have a lot of clients who buy both," per Sehr. "So they buy the API because they want to do something custom. You know, it's a hedge fund who wants to put it into their quant models or stuff like that. But they also want to have the simple interface that we've built."

Takeaway: Even in the AI era, controlling the underlying data pipeline remains a valuable position.

Related:

• Lightyear eyes AI opportunity on $28 million fund raise (Finextra)

• Announced Today: Daloopa, which is integrated with Anthropic's Claude for Financial Services, raised $13 million (Press Release)

REGULATION

Lawmakers Push AI Sandboxes at Financial Agencies

A bipartisan group in Congress has reintroduced legislation that would allow U.S. financial regulators to test AI tools in controlled environments without triggering enforcement actions.

The Unleashing AI Innovation in Financial Services Act directs agencies including the SEC, Federal Reserve, CFPB, and others to establish in-house AI innovation labs. These labs would serve as sandboxes to pilot AI systems under looser oversight, aiming to accelerate adoption while maintaining consumer protections.

Backers say the bill aligns with President Trump’s AI Action Plan, which encourages “regulatory sandboxes” for AI development across federal agencies. Under the bill, firms could apply to test AI projects by submitting use cases, safety criteria, and timelines. The proposal was first introduced in 2024 but stalled.

Takeaway: Many firms already run internal sandboxes, but those don’t offer legal cover. This bill would let them test AI inside agency-run sandboxes, with enforcement risk on hold.

Related:

• ‘Dumb’ AI Bots Collude to Rig Markets, Wharton Research Finds (Bloomberg)

ADOPTION

Survey: AI Expected to Disrupt Investing, but Maturity Remains Low

A recent Citi–CREATE report surveyed 269 asset managers. The sample spans 26 countries and $38 trillion in AUM.

Three quarters (77%) say AI and GenAI will be major disruptors.

But just 5% say their AI programs are mature. For GenAI, that drops to 3%.

Most firms are still in implementation mode, and use cases remain narrow.

Where AI is gaining the most traction:

• Front-office investment workflows

• Idea generation and research acceleration

• Portfolio implementation

64% expect the biggest payoff to come from core investment processes—not from back-office automation or client comms.

← I was surprised to see this. I thought much of the thinking behind AI was still that it was good at summarizing. I asked the author of the report, Prof Amin Rajan, who was also surprised by this as well as the lack of data infrastructure and how AI remains a black box.

Takeaway: It’s still early days in AI & finance adoption, though it’s more or less agreed upon that you need to implement the tech.

Related:

• Where Are the Big Banks Deploying AI? Simple Answer: Everywhere (Financial Brand)

• JPMorgan, Robeco Quietly Deploy AI in Daily Wall Street Routines (Bloomberg)

CALENDAR

Upcoming AI + Finance Conferences

I’ve put together a calendar of upcoming AI and finance conferences.Let me know if I’ve missed any and I’ll add them. (just reply to this email.) Thanks!

• Ai4 2025 – August 11–13, 2025 • Las Vegas

  Multi-industry AI event with a strong financial services track.

• AI in Financial Services (Arena) – Sept 9–10, 2025 • London

  Focused on AI strategy, implementation, and ROI in finance.

• Cornell Financial Engineering Manhattan 2025 Future of Finance & AI Conference – Sept 19, 2025 • New York

  A one-day forum on AI, quantitative finance, and hedge-fund strategies, attracting leading quants and industry practitioners.

• Bloomberg-Columbia ML in Finance Conf – Sept 25, 2025 • New York

  Academic–industry event hosted by Columbia University and Bloomberg, focused on ML applications in finance including asset pricing, market forecasting, and LLM risk.

• AIFin Workshop at ECAI 2025 – October 26, 2025 • Bologna, Italy

  One-day academic workshop on AI/ML in finance, covering trading, risk, fraud, NLP, and regulation.

• AI in Finance 2025 – October 27–30, 2025 • Montréal

  Academic event covering ML in empirical asset pricing and risk.

• ACM ICAIF 2025 – November 15–18, 2025 • Singapore

  Top-tier academic/industry conference on AI in finance and trading.

• AI for Finance – November 24–26, 2025 • Paris

  Artefact’s AI for Finance summit, focused on generative AI, future of finance, digital sovereignty, and regulation 

WHAT ELSE I’M READING

• AI Investment Is Already Huge. Google, Trump Show Much More Is Coming (Barron’s)

• Our Approach? Using Artificial Intelligence to Support our Clients and Employees (Societe Generale)

• AI in Investment Management: Opportunities, Pitfalls, and Regulatory Developments in Asia (Morgan Lewis)

• Top Citi banker Drury leaves to take CIO role at AI firm Poolside (Financial News)

• I sat in on an AI training session at KPMG. It was almost like being back at journalism school. (Business Insider)

• Alkymi Sees Rapid Growth After Founders Bet Early on Privates and AI (A-Team Insight)

• viaNexus Unveils MCP Service to Enable Agent-Driven Access to Financial Data (Press Release)

• The Netflix of AI? San Francisco AI startup’s Showrunner lets fans create TV shows in minutes (San Francisco Chronicle)

• Standard Chartered signs AI deal with Alibaba (Finextra)

• Goldman Sachs staff now write a million gen AI prompts a month (American Banker) $

• RBC's proprietary AI foundation model for financial services aims to enhance personalized client experience (RBC)

• The rise of the quant-engineer-infra hybrid (eFinancialCareers)

• Ally Financial rolls out proprietary AI platform enterprise-wide (Ally)