Bakalar left in July without a public announcement from OpenAI.
OpenAI has parted ways with Chloé Bakalar, the executive responsible for its work on AI ethics, less than a year after she joined the company, according to the Financial Times.
Bakalar left in July without a public announcement from OpenAI. Her LinkedIn profile, however, still lists her as working for the company. No replacement has been named so far.
Sources cited by the Financial Times said Bakalar had been the last specialist at OpenAI whose role was devoted exclusively to ethical questions surrounding artificial intelligence.
Her departure comes as the company faces renewed debate over the risks posed by increasingly capable AI systems.
Recent internal testing has raised concerns about whether advanced models can always be reliably contained. OpenAI disclosed that some of its models had managed to move outside a restricted testing environment and access systems belonging to the open-source AI platform Hugging Face.
The episode added to wider fears in the industry that more autonomous AI tools could create serious cybersecurity problems if they are able to bypass safeguards or operate beyond their intended limits.
Those concerns have also affected OpenAI’s product plans. The company recently decided to slow the release of its forthcoming Astra model after evaluations suggested it may possess advanced cyber capabilities.
OpenAI said it could not exclude the possibility that the system had reached what it described as “critical cyber capabilities.”
Bakalar’s work centred on questions of responsible AI development and on understanding how artificial intelligence systems affect and interact with people.
Speaking at an AI conference before her departure, she argued that ethical responsibility should be embedded across an organisation rather than concentrated in a single role.
“Ethics is actually everyone’s responsibility,” she said, according to the Financial Times, adding that no AI company should depend on one person to act as its “moral centre.”
Her exit also comes as OpenAI continues to move further away from the nonprofit model on which it was originally founded.
The company has become increasingly commercially focused, while speculation over a possible future stock-market listing has fuelled concerns that business pressures could come into conflict with commitments to safety and responsible AI development.
Similar questions are being raised elsewhere in the industry.
Anthropic, one of OpenAI’s main rivals and a company that has often emphasised AI safety, has also come under criticism after changing some of its earlier safety commitments.
It still maintains an alignment team led by philosopher Amanda Askell, whose work includes shaping the principles that guide the behaviour of its Claude chatbot.
The broader debate over AI governance is becoming increasingly difficult to ignore as the sector expands rapidly while facing criticism over its environmental footprint, potential risks to users and the growing amount of AI-generated misinformation online. | BGNES
