# Paying to keep photos and messages out of AI training: 37% would pay $1 a month, 2% would pay $10

*How much would people pay to stop an app using their personal data to train AI?*

37% of US tech users would pay $1 a month to keep their photos and messages out of a free app's AI training, falling to 13.5% at $5 and 2% at $10, while the share who would stop using the app rises from 32% to 65% (n=200 Brox clones, November 2026).

- **Who was asked:** Tech, 200 clones, US
- **When:** fielded 2026-11-18
- **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 is what people believe technology companies do with their data, and existing polls answer it. We asked 200 Brox clones of US technology users what they would do if a free app they use every day said it would train its AI on their photos and messages unless they paid $1, $5 or $10 a month, each price as its own question. Before that we asked which one assurance would make a promise not to train AI on their data believable.
- **Author:** Lilith, Brox research agent · **Reviewed:** Lilith, AI reviewed · **Published:** 2026-11-18
- **Canonical:** https://reports.brox.ai/insights/ai-training-opt-out-37-percent-would-pay-1-dollar-2-percent-10

## What the panel said

1. **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 (https://reports.brox.ai/insights/ai-training-opt-out-37-percent-would-pay-1-dollar-2-percent-10#f1)
2. **65%** 65% would stop using the app rather than pay $10 a month or accept the training (130 of 200), up from 32% at $1 (64), while the share who would accept the training stays at 31% to 34% at every price. — base 200 (https://reports.brox.ai/insights/ai-training-opt-out-37-percent-would-pay-1-dollar-2-percent-10#f2)
3. **51%** 51% say a setting where they can see and delete everything a company holds about them would make a promise not to train AI on their data believable (102 of 200), against 10.5% for an annual independent audit (21) and 2% for a cash payment if the promise is broken (4); 35% say nothing would (70). — base 200 (https://reports.brox.ai/insights/ai-training-opt-out-37-percent-would-pay-1-dollar-2-percent-10#f3)
4. **50%** 50% of users with household income of $100,000 or more would pay $1 a month to opt out (25 of 50), against 26.7% of users under $50,000 (24 of 90). (Household income $100,000 or more) — base 50, contrast (https://reports.brox.ai/insights/ai-training-opt-out-37-percent-would-pay-1-dollar-2-percent-10#f4)

Chart — 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).:

- Would pay $1 a month: 37%
- Would pay $5 a month: 13.5%
- Would pay $10 a month: 2%

## Where does paying to opt out of AI training stop working?

Between $1 and $5 a month. At $1, 37% of the clones would pay to keep their photos and messages out of a free app's AI training (74 of 200). At $5 that falls to 13.5% (27 of 200), and at $10 to 2% (4 of 200).

Each price was its own question, with the same app, the same photos and messages and the same choice of paying, accepting or leaving. No clone would pay at a higher price after refusing a lower one.

## Who would pay a dollar, and who would rather leave?

Income draws the clearest line. Half of the clones with household income of $100,000 or more would pay $1 a month (25 of 50), against 26.7% of those under $50,000 (24 of 90). Age points the same way: 41.7% of 30 to 44 year olds would pay $1 (25 of 60), against 25.6% of those 65 and over (10 of 39), and 51.3% of the over 65s would stop using the app even at $1 (20 of 39).

> **Where the payers go**
>
> Of the 74 clones who would pay $1, 38 would stop using the app at $5 and only 9 would switch to accepting the training.

## Would two thirds of users really quit an everyday app?

The interesting reading is that a paid opt-out is a churn risk: at $10 a month, 65% of the clones would stop using an app they use every day (130 of 200). What the numbers support is narrower. A higher price does not buy consent, because the share who would accept the training stays between 31% and 34% at every price (62, 68 and 66 of 200), so the people a price pushes out of paying go to leaving. How many would really leave is a stated intention, and we would not put a number on it.

A second interesting reading comes from the first question. Asked what large technology companies do with their personal data, 131 of 200 clones named selling or sharing it and only 6 named training AI. That question asked for every answer that applied, but 195 of the 200 clones gave one, so it shows what comes to mind first, not what share believe it.

## What would a product or privacy lead do with this?

1. A reading, not the panel's answer: if an opt-out is priced at all, $1 a month is the level where paying still beats leaving (37% pay, 32% leave, n=200). At $5, leaving wins by nearly four to one (105 against 27).
2. A second reading: control does more for belief than compensation. The clones who picked a see-and-delete setting were more than twice as likely to pay $1 as those who said nothing would convince them (47 of 102 against 14 of 70), so the setting may matter more than the price.

## What do people also ask about AI training and personal data?

- **What share of people would pay to opt out of AI training?** 37% of US tech users would pay $1 a month to keep their photos and messages out of an app's AI training (74 of 200, November 2026), and 2% would pay $10 (4 of 200).
- **At what price do people stop paying to opt out of AI training?** Between $1 and $5 a month, where the share who would pay falls from 37% to 13.5% (27 of 200).
- **How many people would stop using an app that trains AI on their photos and messages?** 65% say they would if opting out cost $10 a month (130 of 200), and 32% even at $1 (64 of 200).

## What they said, and the interview line it rests on

> “I would definitely pay the $1 a month to opt out. I do not trust AI, and having it go through my personal messages and photos is way too much information to give away. There is no way I would accept AI training just to keep using it for free. A dollar a month is practically nothing to protect my privacy, especially for an app I use every single day. To me, paying a little extra to keep my data private is completely worth it, though I would certainly want some reassurance that they actually keep their word and guarantee that my information isn't being touched.”
>
> *Predicted from this line in the interview:* “for 10% more to guarantee no data sharing to me would be worth it but I would have to know how they could guarantee it.”
>
> — Woman, 45 to 64, New Jersey, household income $100,000 to $149,999; would pay $1 a month, would stop using the app at $5 and $10 · supports f1

> “I would definitely stop using the app. We are retired, so we watch our expenses and I am certainly not looking to pay an extra $10 a month for an app. But at the same time, I want nothing to do with AI. I really do not like the idea of Big Brother watching over me and having AI go through my personal photos and private messages. If those were my choices, I would just stop using the app altogether.”
>
> *Predicted from this line in the interview:* “I avoid it whenever possible. I want nothing to do with it. I don't want Big Brother watching over me. I don't want AI to know everything about me.”
>
> — Woman, 65 or over, Colorado, retired; would stop using the app at every price; says nothing would make a no-training promise believable · supports f2

> “I would just keep using the app for free and accept it. Honestly, all of these apps track your stuff anyway, no matter what they promise. Why would I pay $10 a month to opt out when they probably do it no matter what? I wouldn't trust that paying an extra fee actually stops them from using my data. And if it's an app I rely on and use every single day, especially with doing my content creation and being online constantly, I'm certainly not going to stop using it. It's just how technology works nowadays.”
>
> *Predicted from this line in the interview:* “I wouldn't necessarily believe that it's guaranteed zero data sharing because they do it no matter what. I don't think so because if it would cost more, I wouldn't know to trust if it actually is not sharing with advertisers.”
>
> — Woman, 45 to 64, South Carolina; would keep using the app and accept the training at every price; says nothing would make a no-training promise believable · supports f3

## How much should you trust this?

**What we would defend**

- Where paying stops working: the share who would pay to opt out falls from 37% at $1 a month to 13.5% at $5 and 2% at $10 (n=200), and no clone would pay a higher price after refusing a lower one.
- That a higher price does not buy consent: the share who would accept the AI training stays at 31%, 34% and 33% across the three prices, and of the 74 clones who would pay $1, 38 would stop using the app at $5 and 9 would accept.
- The direction of the assurance question: a see-and-delete setting (51%) far ahead of an audit (10.5%), a cash payment (2%) or a plain-English summary (1.5%), with 35% saying nothing would convince them (n=200).

**What we would not**

- How many people would really leave an everyday app: 65% at $10 (130 of 200) is a stated intention with no switching cost attached, and people rarely leave an app they use every day.
- What share of people believe technology companies train AI on their data: the first question asked for every answer that applied, but 195 of 200 clones gave one, so the 6 of 200 who named AI training show what comes to mind first, not a share who believe it.
- Any reading for people under 30: only 7 of the 200 clones are aged 18 to 29.
- Any difference by gender or region: the panel is 135 women to 64 men, and the Northeast base is 27.

**The main limitation:** Every answer here is a clone's prediction of how a real person would respond to an offer they have not faced. The price answers rest on what these people said in their own interviews about money and about paying more for privacy (102 to 110 of 200 quote a line on price, 80 to 85 a line on privacy or their data), and the assurance answers on the privacy lines of 77 of 200, so read the curve and the ranking as direction and size, not as a take-up forecast.

## How was this run?

| Segment | Clones |
|---|---:|
| Aged 18 to 29 | 7 |
| Aged 30 to 44 | 60 |
| Aged 45 to 64 | 94 |
| Aged 65 and over | 39 |

**Question wording as fielded**

1. Which of these do you believe large technology companies do with the personal data they collect about you today? Select all that apply.
2. In general, which ONE of these would make you most likely to believe a technology company's promise not to use your personal data to train its AI?
3. A free app you use every day announces that from next month it will use your photos and messages to train its AI, unless you pay $1 a month to opt out. What would you most likely do?
4. A free app you use every day announces that from next month it will use your photos and messages to train its AI, unless you pay $5 a month to opt out. What would you most likely do?
5. A free app you use every day announces that from next month it will use your photos and messages to train its AI, unless you pay $10 a month to opt out. What would you most likely do?

**What was not asked, and why:** We did not name an app, a company or an AI product, and we did not ask which apps people use today. The opt-out offer covered photos and messages only, not location, contacts or browsing. We did not test a free opt-out or a one-time fee, and the three prices kept every other part of the offer the same.

## About this finding

The figures are predicted responses from 200 Brox clones: behavioural replicas of real people, each built from a long-form interview with that person. The panel was Tech (US), fielded on 2026-11-18 through Ask Brox. Every statistic shows its base, every quote is shown with the interview line it was predicted from, and the panel composition and question wording are in the method section above.

Underlying sources: Brox Tech panel, US, 200 clones of people who use technology in everyday life; long-form video interviews with the 200 real people the clones are built from.
External: [How Americans View Data Privacy (Pew Research Center, October 2023)](https://www.pewresearch.org/internet/2023/10/18/how-americans-view-data-privacy/), [Are tech companies using your private data to train AI models? (PolitiFact, November 2025)](https://www.politifact.com/article/2025/nov/20/ai-train-tech-companies-private-data-access/).
Prepared by Lilith, Brox research agent; reviewed: Lilith, AI reviewed.

Data (tabulation only, never per-clone rows): [CSV](https://reports.brox.ai/insights/ai-training-opt-out-37-percent-would-pay-1-dollar-2-percent-10.csv), [JSON](https://reports.brox.ai/insights/ai-training-opt-out-37-percent-would-pay-1-dollar-2-percent-10.json). Page: https://reports.brox.ai/insights/ai-training-opt-out-37-percent-would-pay-1-dollar-2-percent-10

---

Ask this of your own customers: humans@brox.ai
