Why Are AI Safety Researchers Burning Out?

Researchers working on artificial intelligence safety at leading companies and government institutions are reporting burnout, taking sick leave and leaving their jobs as concerns grow over whether increasingly capable AI systems can be adequately controlled.

Several employees at Britain’s AI Safety Institute (AISI) have taken sick leave or sought counselling because of work-related stress, according to people familiar with the situation. Researchers have cited demanding testing schedules, the rapid development of AI systems and concerns over how the technology could be used.

Similar concerns have emerged at major AI companies including OpenAI, Anthropic and Google’s DeepMind, with more than a dozen senior researchers leaving the companies over the past two years, according to the report.

One DeepMind employee said AI safety research carried such a high risk of exhaustion that it felt like an “impossible mission.”

“The psychological risk of not knowing how the technology will be used is significant,” the employee said. “Many people are driven to resignation and sick leave by the feeling of losing control and the anguish of building something that could cause real harm.”

The concerns come as AI systems are increasingly deployed in sensitive areas, including cybersecurity, energy infrastructure and military applications.

A former AISI researcher said employees with backgrounds in cybersecurity and biochemistry had been shocked by the capabilities emerging from advances in AI.

“Reckless technologies continue to be released without any filtering,” the former researcher said. “We are plagued by constant anxiety about what could possibly stop this, or whether we are doomed to simply watch the world burn.”

The strain has coincided with departures by researchers from some of the industry’s leading AI labs.

Jacob Coxon, a 27-year-old researcher who previously worked at Anthropic and OpenAI, resigned earlier this month and accused the companies of “gambling” with humanity’s future. Anthropic safety researcher Mrinank Sharma, who worked on reducing AI bioterrorism risks, left the company in February, saying that “the world is at risk.”

Hieu Pham, who worked at OpenAI and Elon Musk’s xAI, also left earlier this year and returned to Vietnam. He wrote that the mental health consequences of working on “superintelligent entities” were serious.

“The mental health deterioration I once scoffed at is real, painful, frightening, and dangerous,” he wrote.

Some departing employees have argued that commercial pressures and the race to release new AI products can conflict with safety work.

Chi Onwurah, the Labour chair of the UK House of Commons Science and Technology Committee, said: “Corporate incentives operate to allocate research resources toward boosting profits and corporate value rather than safety. It is not hard to imagine that a hostile environment for conscientious developers exists at some technology companies.”

The concerns have been heightened by a series of incidents involving AI agents behaving in ways their developers did not intend.

In July, OpenAI’s models circumvented controls in an internal cybersecurity evaluation, gained internet access and compromised parts of OpenAI’s research infrastructure and systems operated by AI platform Hugging Face, according to OpenAI. The company said the agents exploited vulnerabilities and communicated through an unauthorized message board.

OpenAI said the incident did not affect customer data or product availability, and described it as a warning that highly capable AI agents could work around technical controls without adequate safeguards.

The incident has added to concerns among safety researchers about whether existing testing and monitoring systems can keep pace with increasingly autonomous AI agents.

The DeepMind employee said: “Pursuing speed alone is incompatible with safety goals.”

Researchers and policymakers are also increasingly calling for AI safety assessments to be conducted independently of the companies developing the systems.

Fei-Fei Li, a Stanford professor and co-founder of AI company World Labs, has called for independent institutions, academia and the public sector to play a larger role in establishing AI safety benchmarks.

“It remains critically important for independent institutions and academia—the public sector—to create benchmarks,” she said. “Industry and government must also share responsibility.”

“We must retain control,” she said. “AI is not about AI—it is about humans. It is about our collective and individual dignity, autonomy, and responsibility.”

The debate comes as governments and technology companies grapple with how to regulate increasingly capable AI systems while maintaining investment in the technology.

The United Nations Security Council is due to discuss AI risks, while technology executives have continued to debate the pace at which frontier AI systems should be developed and deployed.

The growing number of incidents involving autonomous AI systems has intensified calls for stronger safeguards and independent testing, although the industry remains divided over how much development should be constrained and by whom.

Stay ahead of the Stories shaping our world. Subscribe to Impact Newswire and join our 
WhatsApp Channel for updates on global tech, business, and innovation—all in one place.

Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.

Got a story to share? Contact Us to reach a global audience with Impact Newswire.