OpenAI promised last year to launch an opt-out feature to shield content from AI training. Six months after the announcement, there is no more talk.
OpenAI announced Media Manager last May: a portal that would give content creators more control over the use of their content for AI training. The feature would be ready for launch “by 2025. While that doesn’t necessarily mean that OpenAI would launch on Jan. 1, things have remained quiet around Media Manager since the announcement.
OpenAI has received a lot of criticism for how it handled copyright when training its AI models. Copyrighted content was plucked from the Internet for free and for nothing without asking permission or paying a fee. That copyright also exists on the Internet was forgotten for a moment.
read also
OpenAI silences promised opt-out feature for AI training to death
With Media Manager, OpenAI would give content creators more control and better cover itself against possible claims. The tool was announced with much fanfare by OpenAI six months ago. Media Manager would “set a new standard for the AI industry,” according to OpenAI.
No priority
That was the last and only time OpenAI spoke publicly about the tool. Since then, it has been missing from any trace. TechCrunch went in search of an explanation and heard from well-placed sources that OpenAI considers the opt-out feature “not a priority. An ex-employee of the company does not consider himself to remember if the tool was even being worked on. One of the driving forces behind Media Manager left OpenAI in the fall.
For now, content creators must make do with the tools OpenAI offers to shield content from AI training. They can request via a form to have their work excluded from future training sets, and for Web developers, there is a tool to disable OpenAI’s crawlers.
However, those tools are perceived as cumbersome and an explicit opt-out feature is not offered. With some major international news media, OpenAI has entered into a paying agreement.
read also