Meta Platforms has suspended the use of generative artificial intelligence (GenAI) tools in Brazil in response to objections from the Brazilian government and a ban imposed by the National Data Protection Authority (ANPD).
Description
The ANPD raised concerns over Meta’s new privacy policy, which involved using personal data to train GenAI systems in the country [5]. As a result, the ANPD issued a “preventive measure” ordering Meta to stop using user data to train generative AI tools in Brazil [3], citing potential harm to data subjects’ fundamental rights [6]. Meta is currently in negotiations with the ANPD to address their concerns about GenAI and has been given a deadline to prove compliance with the decision.
The ANPD’s decision to suspend Meta’s new privacy policy in Brazil was due to potential violations of the General Data Protection Law (LGPD) [1]. The policy allowed Meta to process user data [1], including children’s data [1], for training its Generative AI without following LGPD principles [1]. ANPD recommended immediate suspension of the policy and processing of personal data for AI training in all Meta products [1]. Meta’s request for a review was rejected for lack of a clear plan to address the issues [1]. Lawyer Cecília Freitas emphasized the importance of complying with LGPD to avoid fines and reputational damage [1]. Companies should base data activities on LGPD legal grounds [1], be transparent with users about data usage [1], and implement special protections for data from children and teens [1].
Brazil is a key market for Meta [5], given its large user base for social network WhatsApp [5]. In June [5], Meta launched an AI-driven ad targeting program for businesses on the popular chat service in Sao Paulo [5]. This move follows Meta’s privacy policy update in May, which allowed the company to train AI models using Brazilian citizens’ public data from Facebook [4], Instagram [4], and Messenger [4]. The ANPD deemed this policy to pose a serious risk to the fundamental rights of Brazilian users [4], leading to the imposition of a daily fine of 50,000 Brazilian Real for non-compliance [4]. The use of personal data to train AI systems without consent has raised privacy concerns globally [6], prompting tech giants like Apple to withhold AI features in regions with stricter data privacy laws [6].
Conclusion
The suspension of GenAI tools in Brazil by Meta Platforms highlights the importance of complying with data protection laws and regulations. Companies must prioritize user privacy, especially when using personal data for AI training [1] [2]. The ANPD’s actions serve as a reminder of the potential consequences of non-compliance, including fines and reputational damage [1]. Moving forward, it is crucial for tech companies to prioritize data protection and transparency to maintain trust with users and regulators.
References
[1] https://lawrope.com/suspension-of-metas-privacy-policy-in-brazil-impacts-on-the-use-of-personal-data-in-generative-ai/
[2] https://www.devdiscourse.com/article/technology/3020824-meta-pauses-ai-tools-in-brazil-amid-privacy-policy-concerns
[3] https://valorinternational.globo.com/business/news/2024/07/18/meta-removes-ai-features-data-protection-agency-analyzes.ghtml
[4] https://uk.pcmag.com/ai/153414/meta-suspends-generative-ai-tools-in-brazil-amid-concern-over-data-training
[5] https://nationaltechnology.co.uk/MetasuspendsuseofgenAItoolsinBrazil.php
[6] https://thehackernews.com/2024/07/meta-halts-ai-use-in-brazil-following.html
