OpenAI announced on Tuesday the formation of a new Safety and Security Committee led by CEO Sam Altman, with support from board members Adam D’Angelo and Nicole Seligman.
Over the next three months, the team will prioritize evaluating and enhancing the company's processes and safeguards.
The new team's main task is "making recommendations on critical safety and security decisions for all OpenAI projects."
OpenAI Board forms Safety and Security Committee, responsible for making recommendations on critical safety and security decisions for all OpenAI projects. https://t.co/tsTybFIl7o
— OpenAI (@OpenAI) May 28, 2024
A future course of action regarding the committee's findings will be decided by OpenAI's board, where Altman, D’Angelo, and Seligman all hold seats.
The fact that the new committee and board have three common leaders raises concerns about whether safety is truly prioritized over product development (and of course, profit).
OpenAI's Recent Safety Issues
Several key figures who recently left the company, including former co-leader of OpenAI's Superalignment team Jan Leike, previously voiced their worries about the tech giant's safety culture.
Although Leike didn't cite specific instances of the company being negligent of safety, he wrote on X after his resignation that "OpenAI must become a safety-first AGI company."
But over the past years, safety culture and processes have taken a backseat to shiny products.
— Jan Leike (@janleike) May 17, 2024
Earlier this month, the company introduced a new voice for ChatGPT named Sky, sparking more controversy due to its similarity to Scarlett Johansson's voice, with the actress slamming OpenAI for using her voice without permission.
Altman explained that there was no intention to replicate her voice, saying the voice actor was cast after failing to reach Johansson.
While the formation of the new safety committee seems to address these issues, the inclusion of major board members in the team calls into question the neutrality of its future decisions.
READ NEXT: Consumers Don't Want AI-Generated Content
Editing by Katherine 'Makkie' Maclang
