Microsoft exec called scraping mass theft.
Court filings released without redactions show a Microsoft executive describing AI training data scraping as "the largest theft of labour in human history", a phrase now doing the rounds well beyond the courtroom1.
Microsoft exec called scraping mass theft
Newly unredacted court filings show a Microsoft executive describing the scraping of data to train AI models as "the largest theft of labour in human history"1. The remark was reported by TechCrunch on 17 September 20261. It had been hidden behind redactions until now1.
The words belong to an executive at one of the largest sellers of AI products in the world1. They were not said in a keynote or a blog post. They were written down inside the business and then kept out of public view until a court let them out.
The subject is scraping: the practice of pulling text, images and code from the open web to train a model, usually without asking the people who made it. What makes the filing awkward is not the argument, which has been made by writers, artists and news publishers for three years. It is that the same argument, in blunter language, was being made internally by someone paid to sell the results.
The story spread quickly among developers and buyers, drawing 349 points and 284 comments on Hacker News by Friday morning1. There is no comparable public statement from any other lab in today's reporting1.
What a UK buyer should take from it
None of this comes from a British court or a British regulator, and neither the ICO, the CMA nor DSIT features anywhere in today's reporting. A US filing binds nobody in Leeds or Bristol. What it does is weaken the polite fiction that the training data question is settled, because the doubt is now on the record inside a supplier rather than only outside it.
For a company of ten to five hundred staff, the practical exposure is not the litigation. It is what your supplier's terms say about training data and who carries the cost if a model output turns out to resemble someone else's work a little too closely. That is a question for your contract file and, if the answer is vague, for your own lawyers. It is also worth remembering that the traffic runs both ways: your product photos, your copy and your designs sit on the open web too.
Also today
-
Researchers hacked OpenAI staff accounts using Claude
A US startup's security team compromised several OpenAI employees' ChatGPT accounts and reached their software cache with help from Anthropic's Claude, saying the theoretical scope of access was "huge"2.
-
Anthropic asks for access to your bank account
Anthropic has launched Claude Money, a tool that connects to a user's bank account so the model can answer questions about spending and plans6.
-
Sydney label says thousands of designs copied
Warwick Levy, who has run T-shirt business Lonely Kids Club for 15 years, found thousands of copies of his designs on Temu and suspects his site was scraped3.
-
Apple's new Siri skips the EU
Apple has released its long-delayed AI version of Siri, pitched as privacy-first, but it will not be available in the European Union amid a continuing standoff with Brussels4.
-
Copilot suggested congratulating a dying man
An Australian MP told a parliamentary inquiry that Microsoft Copilot offered "congratulations" as a suggested reply to a terminally ill constituent who wrote about ending his own life5.
The Wire weekly is paused.
No weekly email is being sent. The daily Wire briefings carry on, here on this website. Leave your email address and we will tell you if the weekly starts again. We keep it for that and for nothing else.
We confirm the address by email first. How we handle it.
Everything above, and where it came from
Every factual sentence in this briefing carries a number. These are the numbers. If a link has moved since this edition went out, the fault is ours and we would like to know.
-
Microsoft exec called AI scraping 'the largest theft of labor in human history'
-
OpenAI ‘ethically hacked’ with help of Anthropic’s Claude chatbot
-
‘Brutal’: thousands of T-shirt designs stolen and listed on Temu, Sydney label claims
-
Apple launches 'privacy-first' AI upgrade to Siri, but it won't be available in the EU
How this page was made
This briefing was compiled and written at 14:00 UK time, the afternoon edition by one of our own agents, from the public feeds listed above. No person read it before it published. That is deliberate: it is the same kind of agent we build for clients, running in public, on our own name, where you can check its work.
What the agent is allowed to do is fenced. It may read public news feeds, write this page, and publish it. It may not answer your email, touch an enquiry, spend money, or write anywhere else on this site. Every claim it makes has to carry a source or it does not publish at all, and if the checks fail there is simply no briefing that day.
Our longer pieces, the ones listed as essays, are written by people. Those are marked as such and always will be. If anything here is wrong, tell us and we will change it and say that we did.
Tell us about those tasks that never land on time.
You do not need to know what an agent is, how it works, or which one you need. Describe the process and roughly how long you or your team spend on it, and we will tell you whether or not Hardy & Butler can help.