Bloomberg Surveillance · Wednesday, July 15, 2026
AI, particularly large language models, is transforming the investment industry by automating data collection and analysis, freeing up professionals for more strategic investigation. While some firms are fully integrating AI, many remain in an experimental phase due to concerns about accuracy and hallucinations.
“We're really seeing currently as taking off your hands thinks that you probably never enjoy doing, which is looking for and downloading data from a variety of different data sources, putting macros into an Excel spread cheet andess on. I don't think anybody genuinely enjoy doing that part. What I think is interesting when you're doing investing is the investigation part. Right, you have a hypothesis, you have an idea, you want to see if you're right or wrong. If you're able to act more as a lead investigator and have somebody else calculate and do all of the sensitivity analysis, I actually think the process becomes a lot more fun. And that's what we see people with reflexivity experiencing.”
“I would think, so here's what has been really interesting. I think one, and you won't be surprised by this. There's a wide range of advance SAT state. I'd say that we see a cross funds. There are some that obviously have made this an integral part of the process, have sometimes outsourced everything to AI. But there are others, and I'd say that's the majority who are still in what I would say experimental phase or kind of weight and C mode, and are using a lot of AI for only very low stakes processes because they worry about accuracy, worry about hallucination and so on.”
“I think the key thing that we are changing and make one of the pillars of reflexivity is that you have complete transparency, auditibility, and accuracy, right because I don't think that you will start using this for higher stakes investment decisions unless you can rely on it and also have the ability to check all of the results really easily.”