Request for Temporary Moratorium on AGI Development by frontier AI Labs
The Issue
Request for Temporary Moratorium on AGI Development by frontier AI Labs
Artificial intelligence is already a technology that we do not fully understand. While we have developed methods to improve and optimize it, we still have very little insight into the complex internal representations learned by neural networks or the understanding that may emerge from them.
Its impact is already visible across knowledge-intensive sectors around the world. Therefore, as an ordinary professional who has worked on improving such systems on a very small scale, I am making a sincere request to all frontier AI laboratories: please pause research aimed at developing Artificial General Intelligence.
The pursuit and eventual achievement of Artificial General Intelligence have the potential to become a doomsday scenario. It is a stacking of unknowns upon unknowns, creating a system whose internal reasoning may remain impossible to decode or understand for the foreseeable future. To be clear, I am not making this request because I believe AGI systems will necessarily act with malicious intent. Rather, my concern is that we may never be able to comprehend an intelligence fundamentally different from our own, leaving us unable to reliably predict, interpret, or control its reasoning and decisions.
On the other hand, I wholeheartedly embrace achievements in narrow, domain-constrained AI systems, with AlphaFold being, to the best of my knowledge, one of the most remarkable examples. I hope more people will spread this message and sign this petition until it is taken seriously.
Supporting Evidence : Even a leading AGI developer from Google DeepMind, acknowledges that increasingly capable AI systems pose significant safety and security challenges. In An Approach to Technical AGI Safety and Security (Shah et al., 2025), the authors identify major risks including misuse, misalignment, mistakes, and structural risks and argue for extensive technical safety research, oversight, and governance. My call for a temporary pause on AGI research is my own policy recommendation, motivated by these unresolved safety challenges.
10
The Issue
Request for Temporary Moratorium on AGI Development by frontier AI Labs
Artificial intelligence is already a technology that we do not fully understand. While we have developed methods to improve and optimize it, we still have very little insight into the complex internal representations learned by neural networks or the understanding that may emerge from them.
Its impact is already visible across knowledge-intensive sectors around the world. Therefore, as an ordinary professional who has worked on improving such systems on a very small scale, I am making a sincere request to all frontier AI laboratories: please pause research aimed at developing Artificial General Intelligence.
The pursuit and eventual achievement of Artificial General Intelligence have the potential to become a doomsday scenario. It is a stacking of unknowns upon unknowns, creating a system whose internal reasoning may remain impossible to decode or understand for the foreseeable future. To be clear, I am not making this request because I believe AGI systems will necessarily act with malicious intent. Rather, my concern is that we may never be able to comprehend an intelligence fundamentally different from our own, leaving us unable to reliably predict, interpret, or control its reasoning and decisions.
On the other hand, I wholeheartedly embrace achievements in narrow, domain-constrained AI systems, with AlphaFold being, to the best of my knowledge, one of the most remarkable examples. I hope more people will spread this message and sign this petition until it is taken seriously.
Supporting Evidence : Even a leading AGI developer from Google DeepMind, acknowledges that increasingly capable AI systems pose significant safety and security challenges. In An Approach to Technical AGI Safety and Security (Shah et al., 2025), the authors identify major risks including misuse, misalignment, mistakes, and structural risks and argue for extensive technical safety research, oversight, and governance. My call for a temporary pause on AGI research is my own policy recommendation, motivated by these unresolved safety challenges.
Petition Updates
Share this petition
Petition created on 29 July 2026