Meta shakes its Responsible AI team

Established in 2019, the Responsible AI (RAI) team has been disbanded by Meta, reported by The Information. The team was initially tasked with scrutinizing AI training approaches at Meta, with a focus on ensuring diversity in data. The objective was to pre-empt issues such as moderation challenges on Meta’s platforms. Meta is owners of social media sites such as Facebook, Instagram, WhatsApp.
The problem arose from automated systems on Meta’s social platforms which included a Facebook translation glitch leading to a false arrest, biased image generation in WhatsApp AI sticker creation, and algorithmic issues on Instagram facilitating the discovery of inappropriate content.
This team is redirected toward the advancement of generative artificial intelligence (AI). As per the report, the majority of RAI team members are set to transition to Meta’s generative AI product team, while others will contribute to the development of Meta’s overarching AI infrastructure. Despite this restructuring, Meta remains committed to the principles of responsible AI development, emphasizing accountability, transparency, safety, and privacy. The company has consistently asserted its dedication to cultivating AI responsibly, prominently featuring its “pillars of responsible AI” on a dedicated webpage.
Jon Carvill, a representative for Meta, affirmed the company’s ongoing commitment to prioritizing and investing in the safe and responsible development of AI. He clarified that although the RAI team is undergoing a restructuring, its members will persist in supporting broader initiatives across Meta related to responsible AI development and usage.
The restructuring at Meta aligns with broader global trends in response to the rapid development of artificial intelligence. Governments worldwide are actively working to establish regulatory frameworks for AI development. The U.S. government has entered agreements with AI companies, and President Biden has directed agencies to formulate AI safety regulations. Simultaneously, the European Union has released its AI principles and is navigating the complex process of passing its AI Act.