I received an email from Mozilla today. I’m including it verbatim, except with all links and names removed.
And I’m writing this from the perspective of someone who has been reading privacy policies for fun (and occasional profit) for nearly 20 years now.
Here is Mozilla’s letter:
Hello,
If you like mysteries, try this spine-chilling read:
Microsoft Services Agreement (in effect as of 30 September 2023)
We read it—along with four lawyers and three privacy experts—and none of us could tell if Microsoft plans on using your personal data to train its AI models.
Before the service agreement went into effect, we asked Microsoft a simple question: Are you using our personal data to train AI?
Microsoft has still not responded to our question, even though this service agreement is now in effect. That’s why we urgently need your help to pressure Microsoft for answers.
Can you add your name to our petition asking Microsoft to say if they’re using our personal data to train AI?
This new agreement covers around 130 Microsoft products—including Office, Skype, Outlook, Teams, Minecraft, and Xbox—and mentions things like audio, video, chat, and attachments. With that much personal data at stake, it’s especially troubling for Microsoft to be so ambiguous about its usage.
With last week’s government actions on AI around the world – including the U.S. government looking more closely at how big tech companies manage personal data and AI – we have an opportunity to get a response.
We don’t know if Microsoft is already collecting and using our personal data, which is why it’s especially important for us to act now. We hope we can count on you to join the tens of thousands of other people who have taken action.
Sign our petition asking Microsoft to say if they’re using our personal data to train AI.
Thank you for all you do for the internet,
I am more than a little bemused by this line:
“We read it—along with four lawyers and three privacy experts—and none of us could tell if Microsoft plans on using your personal data to train its AI models.”
So, I figured I’d give Mozilla a hand.
- Open up the Microsoft Services page.
- Using Firefox, hit ctrl-f.
- Search for “improv” – this will bring up any reference to either improv comedy (less useful, admittedly) or product improvement.
And lo and behold, the first result we get back is this line: “To the extent necessary to provide the Services to you and others, to protect you and the Services, and to improve Microsoft products and services, you grant to Microsoft a worldwide and royalty-free intellectual property license to use Your Content…”
Microsoft offers AI-based products and services. Microsoft claims a license to use customer content to improve products and services. Therefore, Microsoft has the right to use customer content to train AI models as part of product and service improvement. This is clear, and I’m not sure why Mozilla is pretending otherwise.
Three better questions to ask:
- What precise data are used to train AI models, and what data are off limits for use in training models and other machine learning algorithms?
- Why doesn’t Microsoft include these specific commitments in their terms of service?
- What data have been, and/or continue to be shared by Microsoft with OpenAI as part of their ongoing partnership?
And a final question/plea for Mozilla: please, step up your game. You have an incredible platform to inform people of real risks. Why play coy and use situations that require actual advocacy as lukewarm efforts to drum up fundraising and signatures on a petition? We need better.