More than 20 leading artificial intelligence researchers have warned about a possible acceleration in AI Progress.

The researchers published a paper on September 28, 2026. It examines whether automated AI research could trigger an “intelligence explosion.”

The paper was published by the Cambridge Programme on AI Science & Policy. It explores how AI systems could increasingly take part in developing more advanced AI systems.

The researchers describe a process known as recursive self-improvement. Under this scenario, AI systems could identify research limitations and help develop improved successors.

Such a feedback loop could potentially accelerate AI Progress dramatically. The paper discusses a scenario where years of advances could be compressed into months or less.

The paper includes several prominent names from the AI field. Its authors include Geoffrey Hinton, Yoshua Bengio, Jakub Pachocki and Eric Horvitz.

Anthropic co-founder Jack Clark is also among the authors. Researchers from several leading AI organizations contributed to the work.

The researchers point to growing automation within AI research and development. They say AI systems are already handling increasing portions of technical work.

According to the paper, Anthropic reported major growth in AI-generated code. The company said AI systems produced more than 80% of its code by May 2026.

Anthropic also reported growth in autonomous research work. The share of AI R&D completed with high-level human supervision reached 26% in August.

That figure stood at 1% five months earlier, according to the researchers.

The paper also discusses developments at OpenAI. It notes the company has reported progress toward increasingly automated AI research.

The researchers suggest that AI could potentially automate most AI R&D within several years. However, they stress that the timing remains uncertain.

The authors also acknowledge several factors could slow the process. Computing limits, automation challenges and diminishing returns could restrict rapid improvement.

Despite these uncertainties, the researchers say preparation is important. They argue that governments could have limited time to respond if acceleration begins.

The paper calls for greater government oversight of frontier AI development. It recommends improved reporting on the extent of automated AI research.

The researchers also propose independent evaluations and stronger safety testing. They further suggest mechanisms that could allow risky AI research to be paused.

Other recommendations include monitoring high-stakes AI experiments and improving incident reporting. The paper also discusses emergency planning for major AI-related risks.

The researchers say automated AI Progress could bring major benefits as well. Faster research could potentially accelerate scientific and technological breakthroughs.

However, they warn that rapid capability growth could create serious challenges. Society and governments may struggle to adapt if developments move too quickly.

The paper also raises concerns about human oversight. Researchers warn that humans could become less involved in developing increasingly capable systems.

Geoffrey Hinton and Yoshua Bengio have previously raised concerns about advanced AI risks. Their participation adds prominence to the latest research warning.

The researchers do not claim that an intelligence explosion is certain. Instead, they highlight it as a possibility requiring further study and preparation.

Their central concern is the speed of potential change. If automated AI research creates a strong feedback loop, traditional oversight systems may struggle to keep pace.

The paper therefore calls for governments to improve visibility and preparedness. It argues that decisions made before rapid acceleration could become more difficult afterward.

The report concludes that the potential for accelerated AI Progress deserves urgent attention. It also stresses that substantial uncertainty remains about whether an intelligence explosion will occur.

In other news read more about: Mazraoui Injury Raises Concerns Ahead of Key Fixtures