top of page
Search

Use of AI in Investing

Writer: Eddie Perkin
Eddie Perkin
7 days ago
2 min read

How do professional equity investors use artificial intelligence in their investment processes? This is a topic I have been digging into. There is a wide range of use cases and adoption levels.


Morningstar’s “AI in Active Fund Management: The State of Adoption in 2026” report identified fewer than 20 firms that have meaningfully integrated AI in their processes, and those are mostly quant firms.

Based on my conversations, I see five buckets where investors are using AI:

 

1)      Plain vanilla productivity. Sorting emails, generating client reports, preparing investment presentations, taking meeting notes, drafting correspondence.

 

2)      Investment research. Screening companies, gathering data, building models, synthesizing news flow, summarizing earnings calls, tracking sentiment, and generating questions for management.

 

3)      Quality control and validation. Checking models for errors, identifying inconsistent assumptions, and flagging discrepancies in investment reports.

 

4)      Red teaming an investment thesis. In my book, I describe this as “prompt[ing] AI to be your biggest critic.” Challenging assumptions, identifying weaknesses, finding contradictory evidence, and building the strongest possible argument against an investment thesis.

 

5)      Autonomous investment decision-making. Using AI to select securities, construct portfolios, determine position sizes, manage risk, and execute trades. Largely limited to quantitative strategies with human oversight at this stage.

 

But using AI in one’s process and deriving alpha from it are two different things.


Third-party aggregator services like AlphaSense that pull broker research, build models, and generate questions for management flatten the competitive landscape. They help analysts and portfolio managers get up the learning curve quickly, but if competitors are using those same tools, they are unlikely to be a durable source of edge. In fact, groupthink may already be setting in. One equity analyst told me that she has been in group meetings with several hedge fund analysts who each pulled out the same list of management questions.


Likewise for a general-purpose LLM. In theory, you can come up with unique prompts for Claude or ChatGPT, but the underlying models use widely available data, and their outputs can be unreliable.


I believe the opportunity for fundamental investors is to use AI in ways that others are not. Firms with long histories of internal data may be able to use it to develop an edge. These sources include internal research notes, models, trading data, and the like. Alpha could come from using these data sets in unique ways, such as mining your own models, research reports, and trade decisions to identify biases and tendencies.


What other AI use cases for investors have you seen?


 
 
 

Recent Posts

See All
Determinism vs. The Multiverse

The Washington Commanders turned a likely win into a tie by letting the clock run out. It was the most prominent of several heavily debated decision processes from this past weekend's slate of NFL gam

 
 
 

Comments


  • LinkedIn

© 2026 by Eddie Perkin. All rights reserved.

bottom of page