How to Protect Your Data from Meta’s AI Training
Hundreds of thousands of Instagram users, including celebrities like Julianne Moore and Ashley Tisdale, recently shared a post claiming to revoke Meta’s right to use their data for artificial intelligence training. The post states, “Goodbye Meta AI,” and asserts that users do not give permission for Meta or anyone else to utilize their personal data, profile information, or photos. However, this viral Instagram story is actually ineffective and outdated, as confirmed by Meta sources to Sky News. In fact, fact-checkers have flagged the post as “false information” on Instagram.
Meta’s AI Training Plans
Meta, the parent company of platforms like Facebook and Instagram, is gearing up to use public posts and data from UK accounts to train its new artificial intelligence tool, Meta AI. Initially scheduled to roll out in June, the implementation of these new rules was delayed to accommodate changes requested by the Information Commissioner’s Office (ICO).
Easy Steps to Object
Despite the confusion surrounding the viral post, there is a straightforward way for users to object to Meta’s data usage. Meta has promised to honor users’ objections, making the process even simpler, more prominent, and easier to find. In the coming days, Facebook and Instagram users will receive notifications explaining the changes and how to access the objection form. If users have already objected through the previous form, they will not receive a notification as Meta will automatically exclude them from the training data.
Simplified Objection Process
The objection form has been simplified to require only three clicks and an email address to revoke access to personal data. Any non-public information on an account, such as private posts and messages, will already be excluded from the data used for training Meta AI. Additionally, private accounts and accounts belonging to individuals under 18 will also be excluded. Users can also proactively object by visiting Meta’s “privacy center” in their settings and clicking on the “object” hyperlink.
Meta’s Stance on EU AI Pact
While Meta has adjusted its AI policy for UK accounts, it has not yet joined the new pact for companies using AI in the European Union. Earlier this year, the EU passed the EU AI Act, a set of laws that regulate AI usage and categorize different AI applications based on societal risk. Companies operating in the EU, including Meta, are required to comply with these rules, which are gradually being enforced.
EU AI Pact Initiative
In the absence of full implementation of the EU AI Act, an interim set of guidelines known as the EU AI Pact has been introduced. This pact aims to encourage businesses to adhere to the new laws without delay. Companies signing the pact commit to governance strategies aligned with the AI Act, identification of high-risk AI systems, and promotion of AI literacy among employees. While over a hundred companies have pledged their support, Meta has yet to join the initiative.
Meta’s Compliance Efforts
In response to inquiries about joining the EU AI Pact, Meta emphasized its focus on compliance with the AI Act at present. However, the company did not rule out the possibility of joining the pact in the future. Meta expressed optimism about AI’s potential to drive innovation in Europe and foster competition, highlighting the importance of seizing the opportunity presented by AI advancements.
Conclusion
As Meta prepares to implement its AI training using public data from UK accounts, users have the option to object to the use of their personal information. By following the simplified objection process outlined by Meta, individuals can protect their data privacy and ensure exclusion from the training data for Meta AI. While Meta’s compliance with the EU AI Act remains a topic of discussion, the company’s stance on joining the EU AI Pact reflects its commitment to navigating the evolving landscape of AI regulation.