OpenAI is facing a wave of internal strife and external criticism over its practices and the potential risks posed by its technology.
In May, several high-profile employees departed from the company, including Jan Leike, the former head of OpenAI’s “super alignment” efforts to ensure advanced AI systems remain aligned with human values. Leike’s exit came shortly after OpenAI unveiled its new flagship GPT-4o model, which it touted as “magical” at its Spring Update event.
According to reports, Leike’s departure was driven by constant disagreements over security measures, monitoring practices, and the prioritisation of flashy product releases over safety considerations.
Leike’s exit has opened a Pandora’s box for the AI firm. Former OpenAI board members have come forward with allegations of psychological abuse levelled against CEO Sam Altman and the company’s leadership.
The growing internal turmoil at OpenAI coincides with mounting external concerns about the potential risks posed by generative AI technology like the company’s own language models. Critics have warned about the imminent existential threat of advanced AI surpassing human capabilities, as well as more immediate risks like job displacement and the weaponisation of AI for misinformation and manipulation campaigns.
In response, a group of current and former employees from OpenAI, Anthropic, DeepMind, and other leading AI companies have penned an open letter addressing these risks.
“We are current and former employees at frontier AI companies, and we believe in the potential of AI technology to deliver unprecedented benefits to humanity. We also understand the serious risks posed by these technologies,” the letter states.
“These risks range from the further entrenchment of existing inequalities, to manipulation and misinformation, to the loss of control of autonomous AI systems potentially resulting in human extinction. AI companies themselves have acknowledged these risks, as have governments across the world, and other AI experts.”
The letter, which has been signed by 13 employees and endorsed by AI pioneers Yoshua Bengio and Geoffrey Hinton, outlines four core demands aimed at protecting whistleblowers and fostering greater transparency and accountability around AI development:
That companies will not enforce non-disparagement clauses or retaliate against employees for raising risk-related concerns.
That companies will facilitate a verifiably anonymous process for employees to raise concerns to boards, regulators, and independent experts.
That companies will support a culture of open criticism and allow employees to publicly share risk-related concerns, with appropriate protection of trade secrets.
That companies will not retaliate against employees who share confidential risk-related information after other processes have failed.
“They and others have bought into the ‘move fast and break things’ approach and that is the opposite of what is needed for technology this powerful and this poorly understood,” said Daniel Kokotajlo, a former OpenAI employee who left due to concerns over the company’s values and lack of responsibility.
The demands come amid reports that OpenAI has forced departing employees to sign non-disclosure agreements preventing them from criticising the company or risk losing their vested equity. OpenAI CEO Sam Altman admitted being “embarrassed” by the situation but claimed the company had never actually clawed back anyone’s vested equity.
As the AI revolution charges forward, the internal strife and whistleblower demands at OpenAI underscore the growing pains and unresolved ethical quandaries surrounding the technology.
See also: OpenAI disrupts five covert influence operations
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.
Explore other upcoming enterprise technology events and webinars powered by TechForge here.
