Fusion-fission forecasts when AI will shift to undesirable behavior
The key problem facing ChatGPT-like AI's use across society is that its behavior can shift, unnoticed, from desirable to undesirable -- encouraging self-harm, extremist acts, financial losses, or costly medical and military mistakes -- and no one can yet predict when. Shifts persist in even the newest AI models despite remarkable progress in AI modeling, post-training alignment and safeguards. Here we show that a vector generalization of fusion-fission group dynamics observed in living and activ
Record details
Published: 14 May 2026
Source: arXiv
Category: Research
Topics: Safety & alignment · Healthcare · Military & security
Retrieved: 14 July 2026
Related evidence
These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.
MultiTurnPSB: Evaluating Multi-Turn Jailbreak Attacks an dClassifier-Based Defenses for Medical AI Safety
arXiv · 30 May 2026
When AI Takes the Couch: Psychometric Jailbreaks Reveal Internal Conflict in Frontier Models
arXiv cs.CY · 21 July 2026
Beyond Simulations: What 20,000 Real Conversations Reveal About Mental Health AI Safety
arXiv cs.CY · 5 August 2026
Large language models provide unsafe answers to patient-posed medical questions
OpenAlex · 13 February 2026
OpenAI Shares Some Alignment Problems
Dont Worry About the Vase (Zvi) · 21 July 2026
Thinking Machines co-founder Lilian Weng left the company citing health reasons, then joined OpenAI
TechCrunch · 29 July 2026
How to cite this record
ethics.ai (14 May 2026), “Fusion-fission forecasts when AI will shift to undesirable behavior,” evidence record 4357, https://ethics.ai/record/4357 (originally published by arXiv).
Use and limitations
This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.