OpenAI Anthropic
On what the policies say, there's nothing between them. Both are graded D.
OpenAI
ChatGPT trains on your content by default on Free, Plus and Pro, with an opt-out that binds new conversations only. Business and API do not train by default.
Verdict
D
- Trains on you by default
- A self-serve opt-out exists
- On by default. An opt-out exists, but you have to go looking for it.
- 5 Oct 2026
Anthropic
Claude.ai trains on your conversations and other inputs by default on Free and Pro, with an opt-out in account settings. Safety-flagged content is used for training even after you opt out.
Verdict
D
- Trains on you by default
- A self-serve opt-out exists
- On by default. An opt-out exists, but you have to go looking for it.
- 5 Oct 2026
OpenAI: every plan
OpenAI · what each plan can refuse · checked 5 Oct 2026
Free, Plus & Pro (ChatGPT consumer)
YES
Self-serve toggle
Not stated
Not stated
Unknown
32 / 50
On by default, with a way out
15 / 33
One switch, but the policy never says which way it ships
0 / 4
No paywall on the control
0 / 4
One kind of data is in scope
2 / 4
The policy does not say whether people read it
2 / 2
Nothing is said about de-identification
Business, Enterprise & API
NO
Nothing to switch off
Not stated
Not stated
Unknown
13 / 50
Training happens only if you opt in
0 / 33
Nothing is on, so nothing needs refusing
0 / 4
No paywall on the control
1 / 4
2 kinds of data are in scope
2 / 4
The policy does not say whether people read it
2 / 2
Nothing is said about de-identification
Free, Plus & Pro (ChatGPT consumer)TRAINS: YES
- Self-serve toggle
- Not stated
- 51 / 100
- Not stated
- Unknown
Business, Enterprise & APITRAINS: NO
- Nothing to switch off
- Not stated
- 18 / 100
- Not stated
- Unknown
Anthropic: every plan
Anthropic · what each plan can refuse · checked 5 Oct 2026
Free & Pro (Claude.ai consumer)
YES
Self-serve toggle
Not stated
Qualified
Unknown
32 / 50
On by default, with a way out
15 / 33
One switch, but the policy never says which way it ships
0 / 4
No paywall on the control
0 / 4
One kind of data is in scope
2 / 4
The policy does not say whether people read it
1 / 2
De-identification is promised with a qualifier
Team & Enterprise (business offerings)
NO
Nothing to switch off
Not stated
Not stated
Unknown
14 / 50
Off by default
0 / 33
Nothing is on, so nothing needs refusing
0 / 4
No paywall on the control
1 / 4
2 kinds of data are in scope
2 / 4
The policy does not say whether people read it
2 / 2
Nothing is said about de-identification
Free & Pro (Claude.ai consumer)TRAINS: YES
- Self-serve toggle
- Not stated
- 50 / 100
- Qualified
- Unknown
Team & Enterprise (business offerings)TRAINS: NO
- Nothing to switch off
- Not stated
- 19 / 100
- Not stated
- Unknown