AI Content Disclaimer

Last update: 17/05/2026
Hi!

If you’ve reached this page, congratulations.
You do care about the details. And so do we.

Since we’re not here to waste anyone’s time, let’s get straight to the point.

  • Are we creating, sharing, relying upon undisclosed AI-Generated content of any kind to communicate with someone or to (pretend that we) create(d) art of any kind? 
*Drumrolls please*
No.



If you don’t care about the details, no need to go further.
Let me take you back to our homepage (if you’re on your phone, click here instead).

  • A “No” is not enough. Do you use AI for anything in any way?

Yes, sometimes.
  • Wait, so you lied when you answered the first question?!

No. Let me explain:

We have not, do not and won’t use 100% AI-Generated Content.
We will never use AI to create anything from scratch and pretend that it hasn’t been created by some kind of AI.
You will never encounter any kind of AI-Generated content anywhere on our website, in our emails, published content, art of any kind, music, and so on.
And you will certainly never hear anything that is AI-Generated. All of our artists are strongly against AI, especially the undisclosed kind.

In the unlikely scenario that we are unknowingly working with anyone using undisclosed AI in any malicious or even dangerous ways,
send an email to Natha. We will deal with the matter as soon as possible. 

  • Ok. But you do use AI then?

Yes.
We use AI as a tool that helps us taking care of tasks that do not deserve our precious time.
We use LLMs for technical manoeuvres that allows us to focus on the most important aspect of our job: 

Helping our bands. 
We just want to make our life easier, not to become Sam Altman’s BFF.

We are also in a kind of Cold War situation where everyone has started using AI.
Which, paradoxically, forces anyone else to also start using AI.
This is especially true when it comes to emails. More on that below.

So, once again, 
No. We do not use AI-Generated content.

There’s only one (1) situation in which we are actively using AI-Generated content:
In non-human generated notification emails.

As of the time of writing this disclaimer,this is purely restricted to an internal use only.

 Anything else is made with real, human hands and minds.

Period.
If you’re satisfied, let me take you back to our homepage (if you’re on your phone, click here instead).
If you have any question, you can always send us an email to privacy@tempestofnoise.com.

If you want more details, keep reading.

  • I sure do want the details. Which AI?

Claude, from Anthropic.

  • Have you opted out to allow your content (or, worse, our content) to be used to train and improve Anthropic AI models?

We didn’t have to. TON has a Claude for Nonprofits subscription with Anthropic, which is not used for any training or else by default.
From Anthropic’s help page on the subject:

As with all Team and Enterprise plan accounts, data shared with Claude is kept private: by default, we don’t use it to train our models.
Claude maintains SOC 2 Type II, ISO 27001, and HIPAA-readiness, and our connectors don’t access anything that you haven’t explicitly granted them permission to.
See our Privacy Center for additional details.

So, both your data and ours are safe from AI’s training or malicious usage.

  • Ok Ok. But what are you using AI for then?

So, we’ve got numbers and stats.

75% of the time, we do ask Claude to help us with some text-only work:

- Grammar check 
- Typo check
- Translation (in any context)
- Text cleanup
- Etc.

What we don’t do is using AI-generated text in any of our publication, (human sent) emails, etc.
Basically, we don’t use that anywhere. Period.

Besides that, Claude is also at the heart of our email system.

We’re working daily with Shortwave, which helps us at managing the horrifying quantity of emails we’re dealing with.

Basically, Claude helps us not to miss anything important, to clean up the mess in our mailboxes.
It helps us at finding the right information faster, categorising emails in the right folders or doing deep searches in our busy mailboxes*.
Finally, as we previously said, Claude does the final check before pressing “Send”.

We have decided not to use AI-generated quick responses.
So, yep, our “Yes.”, “👍🏻”, “Sounds good!” and co are written by hands.

To summarise, everything that we do at TON has at some point to rely on an email thread.
AI does a tremendous job in that area. And we are fine with it.

*We checked, the biggest mailbox we have has more than ten thousands emails.

  • You said 90%. What about the 10% left?

See? Like us, you also do care about the details :)

10% of the time, we’re doing some deep researches with Claude Opus on multiple subjects (incl. double checking a contract for us).
10% of the time, we’re coding some new tools or improving new ones.
3% of the time, we’re asking Claude misc. technical questions.
Finally, the last 2% are probably wasted in instinctively saying “Thank you” to a computer.

  • Alright. So you truly never ever use AI-Generated content anywhere? 

We literally asked Claude to answer that question for us:

1. You write a draft, I edit it.
2. Research and synthesis.
3. Surgical technical work.

The closest things to "100% AI-generated content" I can find are coding scripts and the memos/outlines inside research reports — and even those are constrained by your spec.

What I don't see in our history: you asking me to write a band biography from nothing, draft a press release without prior copy, generate a complete website, invent marketing copy, write a song, ghostwrite a social post you haven't already drafted, or fabricate a pitch deck. You also turned down the image-editing request on principle (which is a different kind of "create from scratch" refusal, but it tells me something about how you think about origin and authorship).
There’s only one situation where AI-Generated content may happen: inside of non-human generated tech emails.
And that one is for internal use only anyway.

———————————

If you think that we missed mentioning something, don’t understand anything that we mentioned, or have any doubts about what we said,
send us an email at privacy@tempestofnoise.com.

We will make sure that you get the answer you’re looking for.

Long story short: