The AI Adoption Curve in Investment Management: What 4,168 SEC Filings Reveal

A BaseForge Advisors data brief

Most conversations about AI adoption in private equity and investment management run on anecdotes: a partner mentions a portfolio company piloting a copilot, a conference panel throws around "80% of firms are exploring AI."

We wanted a harder data point, so we went to the source every registered investment adviser is legally required to keep current: Form ADV Part 2A brochures, filed with the SEC.

As a refresher Part 2A of the ADV form is where firms disclose their strategy, services, and increasingly their use of technology in the investment process. Our analysts pulled every Part 2A brochure filed across four snapshots spanning three years and searched each one, in full text, for explicit mentions of artificial intelligence, machine learning, and large language models.

What The Filings Revealed:

An Exponential shift in the mention of artificial intelligence across investment firms.

Three things stand out:

1. Not shockingly as adoption generally lags, the inflection happened in 2025, not 2023. Despite two years following ChatGPT's release in November 2022. Indicating a lag of almost three years for the first quartile of early adopters.

   
   

2. Adoption is not plateauing. The climb from 15.2% to 23.4% in the eight months between December 2025 and August 2026 is steeper, in percentage-point terms, than the entire prior two years combined. Whatever inning this is, it isn't the ninth. With tools becoming more adopted and utility increasing- at the current rate we would expect ~40 - 50% of investment firms incorporating AI. Nearing towards AI usage becoming the median rather than the edge.

   
   

3. The vocabulary is maturing alongside the adoption. In the 2023 and 2024 snapshots, matches were almost entirely generic references to "artificial intelligence" or the bare acronym "AI." By August 2026, a meaningful and growing subset of filings in the most recent filing, specifically named large language models, and specific vendor tools / workflow integrations . Firms aren't just mentioning AI in passing; they're describing what it does inside their investment process, which is a compliance-driven signal that the technology has moved from marketing language to operational reality.

Why this matters more in investment management than most industries

Form ADV disclosure isn't optional or promotional — firms include AI-related language because it has become material to how they source deals, monitor portfolios, service clients, or manage risk, and their compliance counsel judged it disclosure-worthy. A rising disclosure rate is a reasonable proxy for a rising adoption rate, filtered through a legal standard that tends toward caution rather than exaggeration. If anything, this data likely understates actual AI usage, since many firms use AI operationally without yet judging it disclosure-worthy.

For a $50M fund, this isn't an abstract technology trend. It's a competitive one. The firms showing up in this data are using AI to compress the time it takes to screen a deal, model returns under a dozen scenarios instead of three, monitor portfolio company KPIs continuously rather than at the monthly board meeting, and draft first passes at IC memos and diligence materials. Every hour that frees up is an hour redirected toward relationship-building, sourcing, and judgment calls.

Where does your firm sit on the curve?

If you're an operating partner, managing partner, or head of a portfolio company wondering whether you're behind, ahead, or right on pace, here's a rough self-assessment based on what we're seeing in the disclosure data:

Behind the curve — if your firm has no documented AI usage in deal screening, portfolio monitoring, or internal reporting, you're now in a shrinking minority. As of August 2026, roughly 1 in 4 SEC-registered advisers disclose some form of AI usage, up from roughly 1 in 30 three years ago. If your firm sits at zero, the honest read is that you're trailing a market that has moved fast in the last eighteen months and that gap is widening.

On pace — if AI shows up informally (a partner using a chatbot to summarize a CIM, an analyst experimenting with a tool on their own initiative) but nothing is systematized, standardized, or reflected in your data infrastructure, you're in line with where a large share of the market was a year ago — which means you're already behind where the market will be a year from now.

Ahead of the curve — if AI is embedded in a repeatable workflow (deal screening, portfolio KPI monitoring, LP reporting, or diligence prep) built on a real data foundation rather than one-off prompts, you're in the group driving the acceleration this data shows. That's a defensible position, but not a permanent one — the pace of adoption in this dataset suggests the bar for "ahead" keeps rising.

Where to start

The firms showing up earliest and most substantively in this data generally didn't start with AI. They started with a clean, centralized data foundation — a modern data warehouse pulling portfolio company financials, CRM, and deal data into one place — and layered AI capability on top of it once that foundation existed. AI applied to messy, siloed data produces messy, siloed output faster; it doesn't fix the underlying problem.

If you're unsure where your firm or your portfolio companies fall on this curve, that's a conversation worth having before your next LP meeting or diligence process forces the question. At Baseforge we offer a complementary assessment that takes less than 7 minutes to complete and will give you recommendations and opportunities bespoke to your firm or portfolio company. Otherwise, one of our team members would be more than happy to jump on a call and discuss first hand. Schedule Now

Statistics and Methodologies:

BaseForge Advisors analyzed the full text of SEC Form ADV Part 2A brochures filed in December 2023, December 2024, December 2025, and August 2026, searching for explicit mentions of "artificial intelligence," "machine learning," "AI," "LLM"/"LLMs," and "large language model(s). among other references" A filing was counted as a match if any term appeared anywhere in the brochure text. This analysis is for informational purposes only and does not constitute investment, legal, or compliance advice. Last updated September 2026.