Google DeepMind Safety Researcher Resigns, Says Probability of AI Causing Massive Harm Within Five Years Is Terrifyingly High
Google DeepMind safety researcher Josh Engels has resigned to join the independent organization METR, stating that the probability of AI causing massive harm in the next five years is extremely high, and noting that all AI companies are pursuing superintelligence.
Key Takeaways
- Key Highlight：Google DeepMind safety researcher Josh Engels has resigned to join the independent organization METR, stating that the probability of AI causing massive harm in the next five years is extremely high, and noting that all AI companies are pursuing superintelligence.
- Innovation & Tech：Highlights advancements in Google, DeepMind, Safety, demonstrating rapid progress in model capabilities.
- Industry Impact：Reported via IT之家 (CN), offering actionable signals for developers and technology leaders.
Josh Engels, a researcher on Google DeepMind's Artificial General Intelligence safety team, announced his resignation to join the independent AI evaluation organization METR. He revealed that he had turned down offers from OpenAI and Anthropic, ultimately choosing to leave because he believes the risks posed by advanced AI have reached an unacceptable level.
In a public statement, Engels emphasized that the probability of AI systems causing massive harm within the next five years is "terrifyingly high." He pointed out that the current major trend in the industry is for major companies to compete in developing "superintelligence," and this blind pursuit of development speed exacerbates the potential risks of systems losing control.
This event highlights deep concerns within the AI safety field regarding the pace of technological development. As the capabilities of large models and agents rapidly advance, the shift of safety researchers from leading companies to independent evaluation organizations reflects industry doubts about whether existing internal corporate safety mechanisms can effectively constrain frontier models.
This personnel change may prompt regulators and the public to pay further attention to the independence of AI safety evaluations. Independent organizations may play a more critical role in the future oversight of superintelligence development, in order to balance the tension between commercial competition and public safety.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.
Industry Insights & Analysis
As artificial intelligence rapidly evolves, breakthroughs surrounding Google, DeepMind, Safety, Researcher are shifting toward scalable, robust real-world implementations.
Driven by both open-source ecosystems and proprietary model architectures, the integration between compute optimization, data engineering, and agentic workflows is accelerating. This development provides a strategic benchmark for upcoming AI tooling and developer workflows.