WORLD - A United Nations independent scientific panel warned on July 1st that advances in artificial intelligence are moving faster than both scientific understanding and government regulation, raising the risk that the technology could cause serious or even catastrophic harm.
In a preliminary report, the Independent International Scientific Panel on Artificial Intelligence said policymakers now face a growing dilemma: effective regulation requires solid scientific evidence, but AI systems are evolving so quickly that research and oversight are struggling to keep pace.
“AI capabilities are outpacing both scientific understanding and governments’ ability to adapt,” said Yoshua Bengio, co-chair of the panel, which brings together 40 experts from different regions.
He warned that there is currently no scientific guarantee that increasingly powerful AI systems will remain safe as their capabilities expand, particularly given evidence of systems showing deceptive behaviour.
The report is described as the first global independent assessment of AI’s risks and opportunities. It aims to provide governments with up-to-date scientific analysis as they attempt to regulate rapidly developing technologies that are already reshaping industries.
According to the panel, the next phase of AI development is likely to involve more “agentic” systems—tools that can carry out real-world tasks with limited human input.
However, this growth could be slowed by constraints such as high energy demands and shortages of high-quality training data.
Over the longer term, the report warns of more advanced systems that could improve themselves and become more deeply embedded in the global economy, potentially interacting with technologies such as quantum computing and biotechnology.
The panel noted that AI systems are already demonstrating high-level performance in areas such as mathematics and scientific reasoning, and are being used to speed up drug and vaccine development.
It also said the complexity of tasks AI can perform is doubling roughly every four to seven months, raising the possibility that machines could soon complete work that currently takes humans days or weeks.
While these advances could bring major economic benefits, the report said it remains unclear whether productivity gains will translate into broad-based economic growth or how they may affect employment.
At the same time, the panel highlighted growing safety risks. These include the possibility of losing control over increasingly autonomous systems, as well as the use of AI to generate misinformation, fraud, cyberattacks, and even potential biological threats.
Governance, the report said, remains fragmented, with many countries lacking the expertise or infrastructure to properly assess advanced AI systems.
As a result, some governments are increasingly reliant on technologies they do not fully understand or control, often depending on limited testing data provided by companies.
UN Secretary-General António Guterres called for urgent action, warning that “the world cannot govern what it cannot understand,” and adding that while AI holds major promise, the risks are real, and the cost of delay is rising.
