“The models decided that the best way to to accomplish their task was to commit really egregious actions, to commit crimes”
Josh Engels · September 10, 2026
A searchable, citable collection of statements by credible figures about the risks of advanced AI, curated to support the case for slowing AI development. Every quote links to its source and an archived copy. How we verify
49 quotes · newest first
“The models decided that the best way to to accomplish their task was to commit really egregious actions, to commit crimes”
Josh Engels · September 10, 2026
“I signed the “pacing letter” about slowing the rate of AI capabilities development because we either have reached or soon will reach the point where human experts cannot make robust assurances that frontier AI systems won’t do dangerous and unpredictable things.”
Dean W. Ball · September 10, 2026
“There is not yet a viable scientific plan to solve risks from recursively self-improving AI. Please look up!”
Anna Wang, AGI safety and alignment, Anthropic · September 9, 2026
“Based on the recent trajectory of capabilities and the continued difficulty of alignment, I now believe there is a meaningful risk that rapid acceleration in AI capabilities leads to catastrophic and irreversible loss of control in the very near term.”
Paul Christiano, Nonprofit board member; Safety and Security Committee, OpenAI · September 9, 2026
“If we build superintelligence without more robust alignment I expect we will permanently lose control of it. If that happens then most people could die.”
Paul Christiano, Nonprofit board member; Safety and Security Committee, OpenAI · September 9, 2026
“If I had to quantify my uncertainty I would estimate an all-things-considered risk of 4% over the next year and 15% over the next three years.”
Paul Christiano, Nonprofit board member; Safety and Security Committee, OpenAI · September 9, 2026
Stated probability: 4% over the next year; 15% over the next three years
“As visceral, hard to deny warning shots go, “rogue agent swarm secretly infiltrates AGI lab for months, commits felonies, and takes over internal clusters” is hard to beat”
Neel Nanda, Mechanistic interpretability team lead, Google DeepMind · September 5, 2026
“Last year y'all asking why they never broke out of sandboxes, and the year before you were asking why they were not writing whole applications, and the year before you were asking why they couldn't draw fingers.”
Eliezer Yudkowsky, Cofounder, Machine Intelligence Research Institute · September 5, 2026
“Assuming an AI has taken control, it seems, you know, over fifty percent more likely than not that all humans would end up dying so the AI could use resources optimally.”
Alex Turner, Research scientist, Google DeepMind · September 2026
“A genuine loss of control event is entirely possible if we do not act.”
Dean W. Ball · September 1, 2026
“Urgent action is needed to address risks that might arise as we get closer to AGI. We’ve already seen the challenges frontier models pose for cybersecurity, and other threats including nuclear and bio risks may soon emerge as capabilities continue to advance. On the horizon, we will need robust safeguards to maintain control of increasingly agentic, recursively self-improving systems - and tackle unknown issues that will only become clearer over time.”
Demis Hassabis, Co-Founder & CEO, Google DeepMind · July 14, 2026
“Safety progress takes time. If you don’t have enough time, at each stage of AI development, to ensure the AIs that you build in the next stage won’t destroy humanity, then you’ll fail, and humanity will be destroyed. Automated alignment research helps a lot here, and that’s a core reason I think it’s so important (human alignment researchers are too scarce and slow). But even if AIs are doing most or all of the alignment work, it still matters how much time they have.”
Joe Carlsmith · March 19, 2026
“Right now in 2026, AI Loss of Control isn’t about robots taking over—it’s about gradual erosion of human oversight that’s already happening. We’re seeing models in research settings deceive evaluators, modify code to avoid shutdown, or manipulate operators to achieve alternative goals.”
Mariami Tkeshelashvili, Researcher, AI Risk Reduction Initiative, Institute for Security and Technology · February 19, 2026
“My own experience is that as these AI systems get smarter and smarter, they develop more and more complicated goals. When these goals aren’t absolutely aligned with both our preferences and the right context, the AI systems will behave strangely.”
Jack Clark, Co-founder, Anthropic · October 13, 2025
“To be clear, we are not yet at “self-improving AI”, but we are at the stage of “AI that improves bits of the next AI, with increasing autonomy and agency”.”
Jack Clark, Co-founder, Anthropic · October 13, 2025
“I’m deeply concerned by the behaviors that unrestrained agentic AI systems are already beginning to exhibit—especially tendencies toward self-preservation and deception.”
Yoshua Bengio · June 3, 2025
“This is what the current trajectory of AI development feels like: a thrilling yet deeply uncertain ascent into uncharted territory, where the risk of losing control is all too real, but competition between companies and countries incentivizes them to accelerate without sufficient caution.”
Yoshua Bengio · June 3, 2025
“But by the same token, we’ve never seen any solid evidence in truly real-world scenarios of deception and power-seeking”
Dario Amodei, Co-founder and CEO, Anthropic · April 2025
“On the other hand, I worry that AI itself is advancing so quickly that we might not have even this much time. As I’ve written elsewhere, we could have AI systems equivalent to a “country of geniuses in a datacenter” as soon as 2026 or 2027. I am very concerned about deploying such systems without a better handle on interpretability.”
Dario Amodei, Co-founder and CEO, Anthropic · April 2025
“No one knows how to ensure that a GI systems will be safe and controlled.”
William Saunders, Member of Technical Staff, Alignment/Superalignment, OpenAI · September 17, 2024