A group comprising OpenAI's current and former employees have released a public letter cautioning against the development of artificial intelligence by the company and its competitors, citing significant risks, inadequate oversight, and alleged suppression of dissenting voices within these organizations.
The letter titled, "A Right to Warn about Advanced Artificial Intelligence", published at righttowarn.ai, outlines a range of potential hazards associated with AI development, including the exacerbation of societal inequalities, dissemination of manipulation and misinformation, and the potential loss of control over autonomous AI systems, which could lead to catastrophic outcomes like human extinction. It noted the role of current and former employees as essential in holding AI corporations accountable, particularly in the absence of effective government oversight.
The letter calls for all AI companies, not just OpenAI, to pledge against penalizing employees who raise concerns about company activities. Additionally, it urges these companies to establish 'verifiable' channels for anonymous feedback from workers on company practices. Critically, the signatories argue that conventional whistleblower protections are inadequate, as they typically focus on illegal conduct, whereas the risks associated with AI development may not yet be regulated. Many employees fear reprisals for speaking out, citing historical instances of retaliation within the industry.
Recently, OpenAI underwent significant changes in its approach to safety management. A research group within OpenAI, tasked with assessing and mitigating the long-term risks associated with the company's advanced AI models, was effectively disbanded last month. This dissolution followed the departure of several prominent figures, with the remaining team members integrated into other groups. Subsequently, OpenAI announced the establishment of a Safety and Security Committee, led by CEO Sam Altman and other board members.
OpenAI spokesperson Liz Bourgeois emphasized the company's commitment to providing the most capable and safest AI systems, affirming their adherence to a scientific approach in addressing risks. Bourgeois acknowledged the importance of rigorous debate surrounding AI technology and pledged ongoing engagement with governments, civil society, and other stakeholders worldwide.
The letter's signatories comprise individuals who previously worked on safety and governance at OpenAI, current employees who opted for anonymous endorsement, and researchers presently affiliated with rival AI companies.