Yahoo News Fact Check: Here's Why Posting This Disclaimer Won't Stop Meta from Using Your Posts to Train AI Models
Claim:
Posting a statement on Instagram or Facebook will stop Meta from using users' data to train generative Artificial Intelligence (AI) models.
Rating:
In late May and early June 2024, a disclaimer was spread on social media, allegedly preventing Meta from using users' content for training its Artificial Intelligence (AI) models. This copypasta, a copied-and-pasted text often used to pass along an alleged warning of some sort, went viral in the light of European Facebook and Instagram users receiving a notification about a privacy policy update that would give Meta permission to use their posts to train its AIs.
"I own the copyright to all images and posts submitted to my instagram profile and therefore do not consent to Meta or other companies using them to train generative AI platforms. This includes future AND past posts/stories/threads on my profile," the copy-pasted text read.
(Facebook user SarahSantos.art)
One Instagram post of the copypasta text read:
Meta is stealing from artists without our consent and feeding our work to their AI. 😡 We rely on these platforms to share our art with the world, to make connections and find work in an industry that's already unstable. I'm not sure what our alternative is, but I'm ready for a place that respects artists enough not to steal our art without consent. Seems like @cara_artists is our next best bet.
I own the copyright to all images and posts submitted to my Instagram profile and therefore DO NOT consent to Meta or other companies using them to train generative Al platforms. This includes all FUTURE and PAST posts/stories on my profile. #noaiart #illustration #artistsoninstagram
"No further posts from me until meta decides to stop using all our posts to train their ai. If you are unaware please look into the situation. So long for now!," another Instagram user wrote. The copypasta was even shared by a renowned tennis player, Rafael Nadal. "More than 250 thousand users shared a message on their IG stories refusing to use their posts to train Meta's AI," the fact-checking organization Maldita reported.
To be clear, simply posting a disclaimer does not override a company's terms of service or negate its privacy policies. Because posting such a disclaimer on Instagram or Facebook would not stop Meta from using users' data to train generative AI models, we have rated this claim as "False."
"How is it that people my age who have lived through several iterations of the fake 'I hereby declare Facebook cannot use my pictures' copypasta thing are posting one on instagram regarding AI training now? Guys cmon," one X post on the topic read.
Jeremy Fisher, a stop-motion director and animator, told NBC News that "many of the fellow artists he follows on Instagram have reshared the viral copypasta statement because of AI-related copyright concerns":
"They're not super involved in the AI conversation for the most part," Fisher said. "It seems like they're trying to check boxes and make sure that this is a thing that protects their work. ... It seems very much just like 'OK, I posted that, and I should be fine and good to go.'"
Facebook's terms of service state that "if you share a photo on Facebook, you give us permission to store, copy, and share it with others (again, consistent with your settings) such as Meta Products or service providers that support those products and services." What's more, a September 2023 article with the title "Privacy Matters: Meta's Generative AI read:
Generative AI models take a large amount of data to effectively train, so a combination of sources are used for training, including information that's publicly available online, licensed data and information from Meta's products and services. For publicly available online information, we filtered the dataset to exclude certain websites that commonly share personal information.
"Publicly shared posts from Instagram and Facebook – including photos and text – were part of the data used to train the generative AI models underlying the features we announced at Connect. We didn't train these models using people's private posts," the company stated, underscoring it wasn't using content of users' private messages to train AI models.
"Meta doesn't have an 'opt-out feature' for users who don't want the company to use their content for its AI training," Meta's spokesperson confirmed to WCNC VERIFY.
In June 2024, Meta announced that until June 26, 2024, European users of Facebook and Instagram would have the option to opt out of having their posts used for training Meta's AI models. This opt-out feature was part of Meta's compliance with stricter data protection regulations in the European Union. However, this opt-out option was not available to users in the United States. On June 14, 2024, the Data Protection Commission reported that Meta had paused "its plans to train its large language model using public content shared by adults on Facebook and Instagram" across the European Union and European Economic Area:
The DPC welcomes the decision by Meta to pause its plans to train its large language model using public content shared by adults on Facebook and Instagram across the EU/EEA. This decision followed intensive engagement between the DPC and Meta.
Read 👉 https://t.co/VkMpdqgcmD pic.twitter.com/IzkF4syVm1— Data Protection Commission Ireland (@DPCIreland) June 14, 2024
Meta's updated statement on the topic read:
We're disappointed by the request from the Irish Data Protection Commission (DPC), our lead regulator, on behalf of the European DPAs, to delay training our large language models (LLMs) using public content shared by adults on Facebook and Instagram — particularly since we incorporated regulatory feedback and the European DPAs have been informed since March.
"This is a step backwards for European innovation, competition in AI development and further delays bringing the benefits of AI to people in Europe," it continued, "We remain highly confident that our approach complies with European laws and regulations. AI training is not unique to our services, and we're more transparent than many of our industry counterparts."
To learn more on the topic, we recommend reading this article from our archives. It's also not the first time we have investigated a viral copypasta. For instance, we debunked the viral claim that posting a legal notice on your Facebook wall would protect you from having all your posts and photos "made public" by Facebook. Moreover, in May 2024, we looked at viral posts retelling a an alleged story about the encounters between former U.S. President Herbert Hoover and Polish pianist Ignacy Jan Paderewski.
Sources:
"Compartir este texto en Instagram no evitará que Meta use tus publicaciones para entrenar a su IA · Maldita.es - Periodismo para que no te la cuelen." Maldita.es — Periodismo para que no te la cuelen, 10 June 2024, https://maldita.es/malditatecnologia/20240610/compartir-texto-no-evita-IA-meta/.
Instagram Post Won't Stop Meta from Using Content for AI Training | Wcnc.Com. 9 June 2024, http://web.archive.org/web/20240609162452/https://www.wcnc.com/article/news/verify/social-media/instagram-post-wont-stop-meta-using-your-content-train-ai/536-b6b44df9-539f-46ca-bd9e-d0893509b3f6.
LaMagdeleine, Izz Scott. "Posting This Notice on Your Facebook Account Will Stop Platform from Using Your Photos?" Snopes, 19 May 2024, https://www.snopes.com//fact-check/facebook-account-notice-photos/.
Log in or Sign up to View. https://www.facebook.com/legal/terms. Accessed 17 June 2024.
"No, You Can't Opt Out of Meta Using Your Posts for AI Training." Social Media Today, https://www.socialmediatoday.com/news/meta-posts-for-ai-training-cannot-opt-out/718410/. Accessed 17 June 2024.
"Privacy Matters: Meta's Generative AI Features." Meta, 27 Sept. 2023, https://about.fb.com/news/2023/09/privacy-matters-metas-generative-ai-features/.
Wrona, Aleksandra. "Herbert Hoover Repaid Favor to Polish Pianist Who Helped Him as Student?" Snopes, 30 May 2024, https://www.snopes.com//fact-check/herbert-hoover-pianist-paderewski/.
