Angel and VC investors turned down an AI-run venture fund at every fee tested: 1 of 30 chose it at 2% and 20%, and still only 1 at a quarter of that
In October 2026, 1 of 30 US angel, venture and fund investors on the Brox panel chose a venture fund run entirely by AI over a human-run fund at equal fees of 2% a year plus 20% of profits, and still only that 1 when the AI fund's fees fell to 1% and 10% and then 0.5% and 5%, while 55.7% of the whole panel (n=61) said a five-year record of real returns is what they would most need first and none named lower fees.
- Who was asked
- US angel and VC investors, 61 clones, US
- When
- October 2026 (fielded 18 October 2026)
- How
- Predicted responses from Brox clones, each built from a long-form interview with a real person.
- What changed from the obvious question
- The obvious question asks whether investors trust AI-run funds, which is answered today with anecdotes about early funds and pilots. This study asked the measurable version: hold a human-run venture fund at standard fees, cut the AI-run fund's fees to half and then a quarter, and count who switches; then ask the one thing that would unlock it.
Section 01: What the panel said
- #
1
A venture fund in which an AI picks every startup, with no human able to overrule it, drew 1 of the 30 private-market investors at equal fees of 2% a year plus 20% of profits, and the same 1 at 1% and 10% and at 0.5% and 5%, while a human-run fund at 2% and 20% drew 14, 10 and 11; of the 29 who passed on the AI fund at equal fees, 0 chose it at either lower fee
base 30 · Invest in private companies: 13 angels, 9 VC, private equity or family office professionals, 8 through funds or syndicates
- #
55.7%
Asked for the one thing they would most need before committing to a venture fund where an AI chooses the investments, 55.7% (34 of 61) named a record of real returns over at least five years, 34.4% (21) said nothing would make them commit, 3 named a human partner signing off every investment, 3 a full explanation of how the AI decides and 0 fees well below a human-run fund; among the 30 private-market investors, 23 named the five-year record
base 61 · contrast
- #
19
All 31 of the 61 who do not invest in private companies chose neither fund at every fee, and 19 of the 31 said nothing would make them commit to an AI-run venture fund
base 31 · Do not invest in private companies · contrast
- #
6
The human-run fund did not gain as the AI fund got cheaper: it drew 14 of 30 private-market investors at equal fees, 10 when the AI fund charged half and 11 at a quarter; the 6 who moved between it and neither cited their own portfolios rather than the AI fund's price, and 1 moved back, so this is read as instability, not a fee effect
base 30 · Invest in private companies: 13 angels, 9 VC, private equity or family office professionals, 8 through funds or syndicates · caveat
A venture fund in which an AI picks every startup, with no human able to overrule it, drew 1 of the 30 private-market investors at equal fees of 2% a year plus 20% of profits, and the same 1 at 1% and 10% and at 0.5% and 5%, while a human-run fund at 2% and 20% drew 14, 10 and 11; of the 29 who passed on the AI fund at equal fees, 0 chose it at either lower fee
02Would angel and venture investors back an AI-run venture fund?
Almost none would, at any fee we tested. We asked 61 wealthy US technology and venture investors on the Brox panel where they would commit $250,000: a new venture fund run by experienced human partners at the standard 2% a year plus 20% of profits, or a new fund in which an AI system chooses every startup, with no human able to overrule it. Thirty of them invest in private companies today (13 as angels, 9 professionally at a venture capital or private equity firm or for a family office, 8 through funds or syndicates). At equal fees, 1 of those 30 chose the AI-run fund, 14 chose the human-run fund and 15 would commit to neither.
The other 31 do not invest in private companies, and all 31 chose neither fund at every fee, so the market this page describes is the 30.
That reading is ours. The strictest reading is narrower: no fee we tested moved anyone towards the AI-run fund, and the one investor who chose it is a single case, not a market share.
03Does cutting the fee change their minds?
No. We wrote the choice out twice more, each time with the human-run fund unchanged at 2% and 20%. With the AI-run fund at 1% a year ($2,500) plus 10% of profits, it drew the same 1 of 30. At 0.5% a year ($1,250) plus 5% of profits, a quarter of the standard fee, it still drew that 1. Of the 29 private-market investors who passed on it at equal fees, 0 chose it at either lower fee.
They saw the discount. At the cheapest rung, 16 of those 29 remarked on the lower fee in their reasoning and turned the fund down anyway. 21 of the 29 explained the choice through judging founders, relationships or access to deals, and 19 named the lack of any human able to overrule the system. One investor who once worked at a venture firm put it most sharply: an autonomous fund would mostly win the deals human investors had already passed on.
The human-run fund did not gain from the comparison either. It drew 14 of 30 at equal fees, 10 when the AI fund charged half and 11 when it charged a quarter. The six investors who moved between it and neither gave reasons about their own portfolios, such as illiquidity, concentration or a preference for index funds, rather than the AI fund's price, and one moved back, so we read that as noise on a borderline commitment and not as a cheap AI fund pulling money away.
04What would an AI-run fund have to show first?
A record, not a discount. Asked for the one thing they would most need before committing to a venture fund where an AI chooses the investments, 34 of 61 (55.7%) named a record of real returns over at least five years. 21 of 61 (34.4%) said nothing would make them commit, 19 of them among the 31 who do not invest in private companies. A named human partner who signs off every investment and a full explanation of how the AI decides were each named by 3 of 61. Fees well below those of a human-run fund were named by none.
Among the 30 who invest in private companies the answer is starker: 23 of 30 named the five-year record, and 2 of 30 said nothing would persuade them.
The one investor who chose the AI-run fund at every fee is a professional who co-founded a venture firm built on quantitative, AI-driven allocation, and who called 2% and 20% legacy overhead. For now, the people ready to back an autonomous venture fund look like the people building one. That is 1 of the 9 professionals, so it is an illustration, not a segment.
05What would a fund manager or allocator do with this?
- Do not lead with price. Cutting the fee to a quarter of 2% and 20% won no one among the 30 private-market investors, and none of the 61 named lower fees as what they would need.
- Lead with evidence. A five-year record of real returns is what 34 of 61 asked for. A reading, not the panel's answer: a manager can build that record with AI working inside a human-run fund before asking anyone to back a fully autonomous one.
- Do not expect a sign-off to carry the case on its own: 3 of 61 named a human partner signing off every investment, the same number as a full explanation of how the AI decides.
What private-fund investors pay is a live question this fall: a Securities and Exchange Commission proposal, summarized by Holland & Knight in October 2026, would widen the investors who may be charged performance fees to include accredited investors. This study suggests that for an AI-run venture fund the fee is not what decides the sale.
It sits alongside earlier Brox pages on AI and advice: high-net-worth investors choose a named advisor over an instant AI assistant, and affluent investors choose a human advisor over an app only below a 1% fee.
06What do people also ask about AI-run funds?
- What share of angel and VC investors would invest in an AI-run venture fund? 1 of 30 US angel, venture and fund investors on the Brox panel (October 2026) chose a fully AI-run venture fund over a human-run one at the standard 2% a year and 20% of profits.
- At what fee would investors choose an AI-run fund over a human-run one? None of the fees tested: at 1% and 10%, and at 0.5% and 5%, the AI-run fund still drew 1 of 30 private-market investors, the same one, against a human-run fund at 2% and 20%.
- How many investors want a track record before backing an AI-run fund? 34 of 61 US technology and venture investors on the Brox panel (55.7%, October 2026) named a record of real returns over at least five years as what they would most need, and 0 named lower fees.
Section 07: What they said, and the interview line it rests on
“I would definitely choose Fund B. This completely aligns with how I view the entire future of investing and innovation capital. When we started our firm, the whole premise was to change how capital is allocated by using science, math, and AI rather than traditional gut-feel methods. We are already building fully automated investment strategies ourselves, and the results are incredible. On top of that, traditional funds charging 2% and 20% feel like legacy institutions charging unnecessary overhead, whereas a 0.5% management fee and 5% carry model powered by autonomous AI agents is lean, modernized, and infinitely more compelling.”
Predicted from this line in the interview
“to change, like how innovation capital is allocated to the entrepreneurs. So we had, like, a fresh perspective about how to make that happen. Using science and math and AI. That's how we got started.”
“I would definitely go with Fund A. Having spent five years working in venture and now running an early-stage angel fund on the side, I know firsthand that early-stage investing is fundamentally about people. Picking winners is very founder- and market-driven, and so much of it comes down to intuition, relationship-building, and having the reps to evaluate a founding team. Beyond that, venture is an access game—an AI system with no human override can't build personal rapport with top founders or win an allocation in a competitive round. Even though the 0.5% and 5% fee structure on Fund B sounds cheap on paper, venture returns follow a power law where access and judgment drive the entire return profile. Paying the standard 2% and 20% for experienced human partners who actually know the space and can evaluate founders is well worth it.”
Predicted from this line in the interview
“on making investments, especially at the early stage, very founder and market driven. And a lot of that's like our intuition level, which is given I have a bunch of reps there”
“As someone who builds AI infrastructure for a living, I know better than anyone how easily models can fail, hallucinate, or look incredible in a backtest while falling apart in production. Anyone can pitch an algorithm that sounds brilliant on paper, but in venture—and in investing generally—paper promises and theoretical models mean nothing without empirical proof. I'm very data-driven; when I started investing, what won me over was seeing the hard, long-term historical data. If you want me to put $250,000 of my liquid money into an AI-managed fund, I would need to see a record of real returns over at least five years. Show me that the system has actually survived market cycles, picked winners, and generated real alpha before asking for my capital.”
Predicted from this line in the interview
“Large language models are incredibly powerful but also incredibly unreliable. They hallucinate, they're inconsistent, they're expensive to run at scale.”
Section 08: How much should you trust this?
What we would defend
- No fee tested moved anyone to the AI-run fund: of the 29 private-market investors who passed on it at equal fees, 0 chose it at 1% and 10% or at 0.5% and 5%; each rung was reasoned on its own fee, and 16 of the 29 noted the lower fee at the cheapest rung and declined anyway.
- Proof outranks price as what would unlock an AI-run fund: 34 of 61 named a five-year record of real returns and 0 named lower fees; among the 30 private-market investors, 23 named the record.
- The 31 who do not invest in private companies are outside this market at any fee: all 31 chose neither fund at every rung.
What we would not
- Any claim about hybrid or AI-assisted funds: only a fund in which an AI picks every startup and no human can overrule it was tested.
- 1 of 30 as a market share: it is one clone, the same one at every fee, and another panel could give none or a few.
- That a cheaper AI fund pulls money away from human-run funds: the human fund drew 14, 10 and 11 of 30 because 6 clones moved between it and neither for reasons about their own portfolios, and 1 moved back.
- That fees never matter: the lowest fee tested was 0.5% a year plus 5% of profits; no zero-fee or performance-only structure was tested.
- The figures as real commitments: $250,000 into a new fund is a hypothetical, and the panel exposes no age, gender or wealth data, so no demographic split is possible.
- The catch-22 and the advice to build a record with AI inside a human-run fund are our readings, not answers the panel gave.
The main limitation
These are 61 clones of wealthy US technology and venture investors, 30 of whom invest in private companies, answering a hypothetical $250,000 choice between one human-run and one fully autonomous AI-run venture fund at three fee levels. It shows that no fee tested moved this panel, not the fee at which one would, and the clones' views of AI come from interviews recorded before this study.
Section 09: How was this run?
| Segment | Clones |
|---|---|
| Invest directly in startups as angels | 13 |
| Invest professionally at a VC or private equity firm, or for a family office | 9 |
| Invest in startups only through funds or syndicates | 8 |
| Do not invest in private companies | 31 |
Question wording as fielded
- Which of these best describes how you invest in private companies today?
- Two new venture funds are raising money and you are deciding where to commit $250,000. Fund A is run by experienced human partners and charges 2% of your commitment a year ($5,000) plus 20% of any profits. Fund B uses an AI system that chooses every startup it invests in, with no human able to overrule it, and charges the same 2% a year ($5,000) plus 20% of any profits. Which would you choose?
- Two new venture funds are raising money and you are deciding where to commit $250,000. Fund A is run by experienced human partners and charges 2% of your commitment a year ($5,000) plus 20% of any profits. Fund B uses an AI system that chooses every startup it invests in, with no human able to overrule it, and charges 1% a year ($2,500) plus 10% of any profits. Which would you choose?
- Two new venture funds are raising money and you are deciding where to commit $250,000. Fund A is run by experienced human partners and charges 2% of your commitment a year ($5,000) plus 20% of any profits. Fund B uses an AI system that chooses every startup it invests in, with no human able to overrule it, and charges 0.5% a year ($1,250) plus 5% of any profits. Which would you choose?
- Which ONE of these would you most need before committing money to a venture fund where an AI system chooses the startups it invests in?
What was not asked, and why
No firm, fund or brand was named. The study did not test a hybrid fund in which an AI recommends and people decide, a zero management fee or a performance-only fee, commitment sizes other than $250,000, or an AI-run fund with an existing track record. The human-run fund was held at 2% a year plus 20% of profits on every rung.
What changed from the obvious question
The obvious question asks whether investors trust AI-run funds, which is answered today with anecdotes about early funds and pilots. This study asked the measurable version: hold a human-run venture fund at standard fees, cut the AI-run fund's fees to half and then a quarter, and count who switches; then ask the one thing that would unlock it.
Fielded through Ask Brox on 18 October 2026.
More on AI in investing and due diligence · Pricing and price sensitivity · Wealth & HNWI · HNWI (VC/Tech) (US) panel
Ask this of your own customers: humans@brox.ai
Related insights
All in this topic-
Angel and VC investors drop an AI startup when its margin after AI model costs falls from 70 to 45 cents on the dollar: 12 of 30 would likely invest at 70 cents, 2 at 45, and none at 20
12 Offered $50,000 of an AI startup selling software to businesses, valued at $20 million with $1 million a year in revenue growing three times a year, 12 of the 30 private-market investors rated themselves likely to invest (4 or 5 on a five-point scale) when it keeps 70 cents of every revenue dollar after paying for its AI models, 2 of 30 at 45 cents and 0 of 30 at 20 cents; no clone rated a lower margin higher than a higher one (base 30)
Read -
18 of the 19 US high-net-worth investors with $2M or more chose a named advisor over an instant AI assistant, and 1 of 30 would move money for better AI
18 Forced to choose one, 24 of 30 US high-net-worth investors took a named advisor they could reach within two business days over an AI assistant available instantly at any hour, and appetite runs the wrong way up the wealth curve: 18 of the 19 holding $2m or more took the advisor, while 4 of the 7 below $1m took the AI. (base 19)
Read -
Health plan pharmacy benefits: 56.5% of members judge them on the price per prescription, and 73.5% would give up their local pharmacy for mail delivery to save $5 a month
56.5% 56.5% of health plan members (113 of 200) say what they pay at the pharmacy for each prescription is the part of their drug coverage that matters most, 41.0% (82) say whether their medicines are covered at all, and 1.5% (3) say using the pharmacy they want near home. (base 200)
Read -
Paying to keep photos and messages out of AI training: 37% would pay $1 a month, 2% would pay $10
37% 37% of US tech users would pay $1 a month to keep their photos and messages out of a free app's AI training (74 of 200), falling to 13.5% at $5 (27) and 2% at $10 (4). (base 200)
Read