Meta's Bold AI Push Comes At Significant Cost, Responsible AI Team Shrinks In Major Restructuring

Amidst the rapid evolution of artificial intelligence, Meta Platforms Inc META is grappling with internal dynamics surrounding its Responsible AI (RAI) team.

What Happened: Initially established to ensure the ethical deployment of AI tools, the group has undergone significant changes in its structure and mission.

The journey of the RAI team began in 2020 with a team of 30, which later expanded to nearly 40 members. However, recent internal shifts have seen its strength reduce to an estimated 25, according to Insider.

Originally envisioned as a vanguard in AI safety, the RAI team has now pivoted more towards compliance or as one Meta insider told the outlet, "How ... we make sure we don't break any laws or get sued again."

With AI taking center stage, Meta has been proactive, launching its Llama AI model and introducing generative-AI features in platforms like Facebook Messenger and WhatsApp.

Yet, the company has also witnessed significant internal changes, including mass layoffs and restructuring. CEO Mark Zuckerberg rationalized these moves as efforts toward achieving "efficiency" and a refocused commitment to core tech innovation.

Following the exit of Jerome Pesenti, former vice president of Meta's AI group, the RAI team underwent further changes. It merged with the social impact team, followed by staff reductions.

Also Read: Harvard Expert Says Zuckerberg Is Derailing Facebook: 'I Think The Wealth Went To His Head'

This reshuffling reportedly coincided with leadership transitions, as Mike Schroepfer relinquished the CTO title to Andrew "Boz" Bosworth, who assumed the AI helm and orchestrated the repositioning of RAI and other teams.

Despite facing challenges like "competing interests and competing for resources," the RAI team managed to release over 20 "system cards" that publicly detailed AI-driven recommendation systems on Facebook and Instagram, according to Insider.

However, the time taken to achieve this milestone raised eyebrows among some insiders.

Pesenti highlighted that the shift towards compliance began last year under the leadership of Esteban Arcaute, RAI's technical engineering lead.

While Pesenti told Insider that he found the change "reasonable," he acknowledged it wasn't universally embraced. As AI regulations continue to evolve, the RAI team's foundational mission may regain its significance. However, current sentiments suggest the company might be underestimating the gravity of responsible AI.

Now Read: Meta's 'Year of Efficiency' Impact: Layoffs Extend To Charitable Organization

This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Photo: Shutterstock