# Mozilla rewrote Firefox's Terms of Use after backlash, saying the change wasn't about using data for AI

> In late February 2025 Mozilla gave Firefox its first-ever Terms of Use; a broad data-license clause sparked a backlash and fears the language could feed AI training or ad-targeting. Mozilla denied that intent and, within days, revised then rewrote the Terms to a narrower, scoped license.

- Verification status: verified
- Case type: deployment
- Provider: Mozilla Corporation
- Client: Mozilla Corporation, Web browser / consumer software (named)
- Sector: software / US / cross
- Verified on: 2026-08-24
- Canonical URL: https://theinternetninja.com/stories/mozilla-firefox-terms-of-use-rewritten-after-backlash-not-using-data-for-ai-2025/
- Source: The Internet Ninja (theinternetninja.com), independent verified-proof platform

## Outcomes

| Metric | Before | After |
| --- | --- | --- |
| Firefox Terms of Use content-license clause (Feb 26 → Mar 3, 2025) | "a nonexclusive, royalty-free, worldwide license to use that information to help you navigate, experience, and interact with online content" — the broad grant introduced Feb 26, 2025 | "a nonexclusive, royalty-free, worldwide license for the purpose of doing as you request with the content you input in Firefox" — scoped, and "does not give Mozilla any ownership in that content" (Mar 3, 2025 rewrite) |

## Verification method

Public terms/data-language walk-back after backlash. Every load-bearing quote is byte-tied to Wayback captures of Mozilla's own Firefox Terms of Use page and blog plus two independent newsrooms (The Hacker News, TechCrunch), each re-fetched and re-grepped verbatim on 2026-08-19. Not a client-outcome engagement; no numeric performance outcome is claimed.

## Full case file

## The problem

On February 26, 2025 Mozilla introduced a Terms of Use for Firefox for the first
time, alongside an updated Privacy Notice [source](https://thehackernews.com/2025/03/mozilla-updates-firefox-terms-again.html).
As preserved verbatim on Mozilla's own archived Firefox Terms of Use page that day,
one clause read: "When you upload or input information through Firefox, you hereby
grant us a nonexclusive, royalty-free, worldwide license to use that information to
help you navigate, experience, and interact with online content as you indicate
with your use of Firefox" [source](https://www.mozilla.org/en-US/about/legal/terms/firefox/).
The Hacker News quoted the same clause word for word, and TechCrunch independently
quoted the core grant — "you hereby grant us a nonexclusive, royalty-free, worldwide
license to use that information" [source](https://techcrunch.com/2025/02/28/mozilla-responds-to-backlash-over-new-terms-saying-its-not-using-peoples-data-for-ai/).
Critics read the broad grant as Mozilla claiming sweeping rights over user data and
feared it could permit using Firefox data to train AI or sell it to advertisers — a
fear sharpened because Mozilla had also stepped away from a blanket "we never sell
your data" assurance [source](https://blog.mozilla.org/en/firefox/update-on-terms-of-use/).

## What changed

This is not a product deployment; it is a terms/data-language change and its
reversal, in two distinct steps. Responding to the backlash on February 28, 2025, a
Mozilla spokesperson said the changes "are not driven by a desire by Mozilla to use
people's data for AI or sell it to advertisers" [source](https://techcrunch.com/2025/02/28/mozilla-responds-to-backlash-over-new-terms-saying-its-not-using-peoples-data-for-ai/).
On its blog the same day Mozilla explained why it had dropped the blanket no-sale
language: it had "stepped away from making blanket claims that 'We never sell your
data' … because, in some places, the LEGAL definition of 'sale of data' is broad
and evolving" [source](https://blog.mozilla.org/en/firefox/update-on-terms-of-use/).
It also said it had "removed the reference to the Acceptable Use Policy because it
seems to be causing more confusion than clarity" [source](https://blog.mozilla.org/en/firefox/update-on-terms-of-use/).

## The outcome

Mozilla then narrowed the license. Its blog reframed the grant as "a nonexclusive,
royalty-free, worldwide license for the purpose of doing as you request with the
content you input in Firefox" [source](https://blog.mozilla.org/en/firefox/update-on-terms-of-use/),
and The Hacker News independently quoted the revised clause — "royalty-free,
worldwide license for the purpose of doing as you request with the content you input
in Firefox. This does not give Mozilla any ownership in that content" [source](https://thehackernews.com/2025/03/mozilla-updates-firefox-terms-again.html).
On March 3, 2025 TechCrunch reported Mozilla rewrote the Terms again, describing the
change as one meant to "more clearly reflect the limited scope of how Mozilla
interacts with user data" [source](https://techcrunch.com/2025/03/03/mozilla-rewrites-firefoxs-terms-of-use-after-user-backlash/).

**Scope guard.** This is a wording clarification and a Terms rewrite in response to
backlash — NOT an admission that Mozilla ever used Firefox data to train AI or sold
it. Mozilla's stated position is a denial of that intent; the AI angle here is the
users' fear plus Mozilla's denial, not a confirmed practice. Note too that the
phrase "nonexclusive, royalty-free, worldwide license" appears in *both* the
original and the rewritten clause — what changed is the qualifier ("to use that
information to help you navigate…" became "for the purpose of doing as you request…"),
so each quote above carries enough of the clause to show which version it is.

**On sourcing.** The strongest evidence is first-party (Mozilla's own archived Terms
page and blog), but a first party narrating its own reversal is not independent proof
on its own; the two load-bearing before/after facts are therefore each corroborated
by two mutually independent newsrooms (The Hacker News and TechCrunch), not by
Mozilla alone. The Feb 28 spokesperson denial rests on a single outlet (TechCrunch),
and its attribution varies across newsrooms — so it is treated as non-critical
context, not a load-bearing claim.

## How this was verified

- **Method:** public terms/data-language walk-back after backlash — a documentary
  trail, not a client engagement. Each load-bearing quote was re-fetched from its
  specific Wayback capture and re-grepped verbatim in-capture.
- **Sources:** Mozilla's own Firefox Terms of Use page (Feb 26, 2025) and its
  "update on our Terms of Use" blog (first-party origins), plus two independent
  newsrooms, The Hacker News and TechCrunch (Feb 28 and Mar 3 pieces).
- **Independence:** the two before/after facts each bind to The Hacker News +
  TechCrunch as two independents; Mozilla's own pages are the origin and are not
  counted among the two independents.
- **Date verified:** 2026-08-19.
- **Status:** PENDING — publish-ready at its honest band, awaiting the owner's
  sign-off; not `verified`, no green badge. No numeric performance outcome is
  claimed.

## Sources

1. **The Hacker News** · "Mozilla Updates Firefox Terms Again After Backlash Over Broad Data License Language" · Mar 1, 2025 · https://thehackernews.com/2025/03/mozilla-updates-firefox-terms-again.html — **Tier 2** (independent press)
2. **Mozilla** · "Firefox Terms of Use" (archived Feb 26, 2025) · Feb 26, 2025 · https://www.mozilla.org/en-US/about/legal/terms/firefox/ — **Tier 1** (first-party origin)
3. **Mozilla** · "An update on our Terms of Use" (Mozilla blog) · Feb 28, 2025 · https://blog.mozilla.org/en/firefox/update-on-terms-of-use/ — **Tier 1** (first-party origin)
4. **TechCrunch** · "Mozilla responds to backlash over new terms, saying it's not using people's data for AI" · Feb 28, 2025 · https://techcrunch.com/2025/02/28/mozilla-responds-to-backlash-over-new-terms-saying-its-not-using-peoples-data-for-ai/ — **Tier 2** (independent press)
5. **TechCrunch** · "Mozilla rewrites Firefox's Terms of Use after user backlash" · Mar 3, 2025 · https://techcrunch.com/2025/03/03/mozilla-rewrites-firefoxs-terms-of-use-after-user-backlash/ — **Tier 2** (independent press)

## Related case files

- [Dropbox walks back its third-party-AI toggle](/stories/dropbox-third-party-ai-openai-toggle-backlash-data-not-used-to-train-models-2023/) — Dropbox walks back a third-party-AI toggle after users feared their files trained OpenAI models.
- [Zoom reverses its AI-training terms on customer content](/stories/zoom-terms-of-service-ai-training-customer-content-reversed-after-backlash-2023/) — Zoom reverses a Terms-of-Service clause on training AI with customer content.
- [Adobe adds an AI-training clarification to its Terms of Use](/stories/adobe-terms-of-use-ai-training-clarification-added-after-backlash-2024/) — Adobe adds a Terms clarification denying it trains AI on user content after a similar scare.
- [WeTransfer removes its AI-training clause](/stories/wetransfer-ai-training-clause-terms-removed-after-backlash-2025/) — WeTransfer removes an AI-training clause from its terms after user backlash.