The parent company of Tumblr and WordPress, Automattic, is striking a deal to sell user data to train artificial intelligence (AI) models.
The deal has been discovered by 404media, with the tech news site getting hold of internal documents that show the impending data sale with companies like Midjourney and OpenAI.
404media alleges that there will be a new setting introduced today that will ”allow users to opt-out of data sharing with third parties, including AI companies.”
How Automattic plan to use AI
Automattic has a statement on their site that is titled “Protecting User Choice” and it goes into detail about the company’s stance on AI.
“AI is rapidly transforming nearly every aspect of our world, including the way we create and consume content,” the statement reads. “At Automattic, we’ve always believed in a free and open web and individual choice. Like other tech companies, we’re closely following these advancements, including how to work with AI companies in a way that respects our users’ preferences.”
The company also state it currently blocks AI crawlers and search engine indexing sites to keep user’s posts out of their eye-line, unless a user has agreed for this content to be public.
Automattic would also openly discuss their work with third parties saying “we are also working directly with select AI companies as long as their plans align with what our community cares about: attribution, opt-outs, and control.”
”Our partnerships will respect all opt-out settings,” the statement on AI would read. “We also plan to take that a step further and regularly update any partners about people who newly opt out and ask that their content be removed from past sources and future training.”
So a deal with AI companies and the specific impacts on user data is yet to be seen, but the correlation between the 404media report and this public statement shows that Automattic is already considering the implications of public-facing information and artificial intelligence.
Why do AI companies want access to user data?
AI companies and their complicated algorithms need a data source to be essentially “trained” on a set or specific topic.
The better the data that feeds these learning models, the more exact the output will be in theory. So for example an AI model that can only access limited data will have a limited authority on a certain topic.
Impeding regulation on AI, which has been a rather lawless environment means that more deals must be struck with the companies that have access to these data sets for the best return of an AI query. This also means the owners of the learning models don’t fall foul of the legal side of data copyright and those who own platforms like Tumblr and WordPress get something back for access to the user data.
Featured Image: DALL-E
