IT·Science News

AI giants back a scraping opt-out, as copyright fights heat up in Australia and New York

OpenAI, Google, Microsoft and Anthropic say they will adopt a new standard that publishers can use to opt out of having their content scraped for AI, The Verge reported on October 30, following an earlier report by Search Engine Roundtable. Instead of setting AI training preferences separately for each company, publishers would send one shared "policy token" signal that applies across all of them.

The same day brought two reminders that the fight over training data is far from settled. In Australia, The Guardian reported that Coalition senators and the recording industry warn that changing copyright law to let AI firms scrape music, books and journalism could breach the constitution and open the country to compensation claims from rights holders worldwide. Constitutional expert Anne Twomey said such scraping could potentially amount to an acquisition of property by the government.

“Extremely alive” to the issue. — Australia's attorney general's department, via The Guardian

In the US, Variety reported that the Human Consent Foundation, a nonprofit backed by Cate Blanchett, and a group of Hollywood companies filed an amicus brief in the New York Times' lawsuit against OpenAI and Microsoft. They argue that Hollywood already has a "proven infrastructure" for licensing content, against the tech companies' position that clearing the rights was not reasonably possible.

How each outlet framed it

• The Verge — a practical fix: one switch for publishers instead of many.
• The Guardian — the legal risk to governments that loosen copyright for AI.
• Variety — the entertainment industry's argument that licensing is workable, not impossible.

What we don't know yet: when the opt-out standard takes effect, whether it covers content already scraped, and whether Australia will go ahead with the copyright change.