Meta pauses plans to train AI using European users' data, bowing to regulatory pressure | TechCrunch

Meta pauses plans to train AI using European users' data, bowing to regulatory pressure | TechCrunch

Meta has it. Confirmed That it will halt plans to start training its AI systems using data from its customers in the EU and the UK.

The move follows pushback from the Irish Data Protection Commission (DPC), Meta's lead regulator in the EU, acting on behalf of data protection authorities across the bloc. UK's Information Commissioner's Office (ICO) Also requested That Meta put his plans on hold until he can address his concerns.

“DPC welcomes Meta's decision to halt its plans to train its large language model using public content shared by adults on Facebook and Instagram in the EU/EEA,” DPC said in a statement. I said. Statement Friday. “This decision follows intensive engagement between the DPC and Meta. The DPC, in cooperation with its fellow EU data protection authorities, will continue to engage with Meta on this issue.

While the meta is already there. Tapping user-generated content to train its AI In tough markets like US, Europe GDPR regulations This has created obstacles for Meta — and other companies — trying to improve their AI systems, including large language models with user-generated training material.

However, Meta started notifying users about one last month. The coming change In its privacy policy, which it says will give it the right to use public content on Facebook and Instagram to train its AI, including comments, interactions with companies, status updates, photos and include captions associated with them. Company argued that he needed to do so. Reflecting “the diverse languages, geographies and cultural contexts of people in Europe”.

The changes were to take effect on June 26, 12 days from now. But Encouraging projects Non-profit privacy activist organization NOYB (“None of Your Business”) to file 11 complaints with EU member states, arguing that Meta is violating various aspects of the GDPR. One of them relates to the issue of opt-in versus opt-out. vis à vis Where personal data processing takes place, consent must first be obtained from users rather than requiring action to opt-out.

Metta, for its part, was relying on a GDPR provision called “legitimate interests” that its actions were in line with the regulations. This is not the first time Metta has used this legal ground in defense. After doing so first To justify the processing of European users for targeted advertising.

It always seemed likely that regulators would at least halt implementation of Meta's planned changes, especially given how difficult the company has made it for customers to “opt out” of using their data. The company said it has sent more than 2 billion notifications informing users of upcoming changes, but unlike other important public messages plastered at the top of users' feeds, e.g. Indicates to go out and vote., these notifications appeared alongside standard user notifications: friends' birthdays, photo tag alerts, group announcements and more. So if one does not check their notifications regularly, it was very easy to miss it.

And those who saw the notification wouldn't automatically know there was a way to object or opt out, because it invited users to click only to find out how Meta would use their information. . There was nothing to suggest there was a choice.

Meta: AI notification
Meta: AI notification
Image credit: Meta

Moreover, users were not technically able to “opt out” of having their data used. Instead, they had to complete an objection form where they set out their arguments as to why they did not want their data to be processed – it was entirely at Meta's discretion whether this request was honored. , though the company said it would respect everyone. Application

Facebook "Objection" Form
Facebook “Objection” Form
Image credit: Meta / screenshot

Although the objection form was linked to the notification itself, anyone who actively looked for the objection form in their account settings was out of business.

On Facebook's website, they had to click on them first. Profile picture top right; kill Settings and privacy; tap Privacy Center; Scroll down and click on Generative AI in the meta Scroll down again after a bunch of links to the Section Title section. More resources. The first link under this section is called “How metagenerative AI uses information to make models.“And they needed to read about 1,100 words before going to a discrete link to the company's “right to object” form. The Facebook mobile app had a similar story.

Link to "Right to object" Form
Link to the “Right to Object” form.
Image credit: Meta / screenshot

Earlier this week, when asked why the process required a user to object rather than opt-in, Meta's policy communications manager Matt Pollard pointed this out to TechCrunch. Current blog postwhich says: “We believe that this legal basis (“legitimate interests”) is the most appropriate balance for processing public data on a scale that respects people's rights, to train AI models. It is necessary.”

To translate this, opting in won't create enough “scale” in terms of people willing to submit their data. So the best way around this was to issue a stand-alone notification among other user notifications. hide the objection form behind half a dozen clicks for those willing to freely “opt out”; And then justify their objection instead of giving them an outright opt-out.

In one Updated blog post On Friday, Meta's director of global engagement for privacy policy, Stefano Fratta, said he was “disappointed” by the request from the DPC.

“This is a step backwards for European innovation, competition in AI development and further delays in bringing the benefits of AI to people in Europe,” Fratta wrote. “We are very confident that our approach complies with European laws and regulations. AI training is not unique to our services, and we are more transparent than many of our industry counterparts.

AI arms race

None of this is new, and the meta is in. AI arms race which has flashed a A giant spotlight on the vast arsenal of data Big Tech is all over us.

earlier this year, Reddit revealed that it had been contracted to make it. North of $200 million over the coming years to license their data to companies Like ChatGPT-maker OpenAI And Google. And these companies already have the latter. Face heavy fines To train its creative AI models to rely on copyrighted news content.

But these efforts also highlight the lengths to which companies will go to ensure they can leverage that data within existing legislative restrictions. “Opt-in” is rarely on the agenda, and the opt-out process is often unnecessarily difficult. Just last month, Someone saw some suspicious words. In the current Slack privacy policy, which suggested it would be able to leverage user data to train its AI systems, users would be able to opt out simply by emailing the company.

And last year, Google Finally gave way to online publishers Opting out of training its models by enabling your website users to inject a piece of code into their sites. OpenAI, for its part, is Building a dedicated tool allowing content creators to opt out of training its creative AI smarts; It should be ready by 2025.

While Meta's efforts to train its AI on public user content in Europe are still on hold, it will likely resurface in another form after consultation with the DPC and ICO — hopefully users With different permission processes.

Stephen Almond, ICO's executive director of regulatory risk, said, “To make the most of generative AI and the opportunities it brings, it is critical that the public trust that their privacy rights are protected. will be respected from the start.” a Statement Friday. “We will continue to monitor the major developers of generative AI, including Meta, to review the security measures they put in place and ensure that the information rights of UK consumers are protected.”

About the Author

Leave a Reply