TheCorporateCounsel.net

The Center for Audit Quality, together with KRC Research, recently published this 24-page institutional investor survey on the use of artificial intelligence in research and analysis for investment decision making.

The survey was conducted in May 2026 and represents views from 100 investors with $500 million or more in assets under management (the appendix has a lot of detail about job titles, type of investor, etc.). It follows a CAQ survey that Dave blogged about earlier this summer on how investors felt about using GenAI in public company audits.

Here are a few key findings:

– Earnings calls are one of the most common “use case” – with 16% using AI all the time to analyze earnings calls transcripts, 45% using it often, 14% using it occasionally, 13% using it rarely, and 13% never. If you don’t have your math hat on, that means 60% of investors said they use AI tools “always” or “often” to analyze earnings calls.

– For company filings like 10-Ks and 10-Qs – 16% of investors are using AI to analyze filings all the time, 25% often, 33% occasionally, 13% rarely and 14% never.

– Investors are also using AI tools on risk assessment scenario analysis (49% always or often), financial modelling and projections (46% always or often), and screening investment opportunities (45% always or often).

The survey gives this additional color on how AI tools are used to review company filings:

– Extracting specific financial or operational metrics – 60%

– Summarizing key sections like the MD&A, risk factors, etc. – 58%

– Comparing filings across similar companies in the same industry – 50%

– Identifying inconsistencies between narrative and financials – 48%

– Comparing filings across time periods for consistency or significant changes – 47%

– Synthesizing insights across multiple disclosures – 46%

– Identifying anomalies or red flags – 45%

– None of the above – 2%

Companies are already using AI to quality-check filings, but this investor-sourced list might help with your prompts.

– Liz Dunshee