A major United Nations scientific panel has released its first global assessment on artificial intelligence. Its report has found artificial intelligence is advancing faster than governments and regulators can keep up. The experts also warn the technology poses growing risks - including cyber threats, disinformation and potentially catastrophic harm.
A major United Nations scientific panel has released its first global assessment on artificial intelligence.
Its report has found artificial intelligence is advancing faster than governments and regulators can keep up.
The experts also warn the technology poses growing risks - including cyber threats, disinformation and potentially catastrophic harm.
“The panel is intended to help the world separate fact from fakes - and science from slop."
That’s António Guterres, unveiling the United Nations’ first independent scientific assessment of artificial intelligence.
The panel was established by the UN General Assembly last year, bringing together 40 experts drawn from fields including computer science, economics, medicine and journalism.
Its task was simple, but urgent - provide governments with an evidence-based understanding of AI as the technology rapidly evolves.
Guterres says the opportunities are enormous.
"Used well, AI could be the most powerful engine for development, speeding the world's progress on everything from health and hunger to learning and climate. But the panel is just as clear-eyed about the harm artificial intelligence can cause. The co-chairs will go through their findings in detail. But first, let me draw a single lesson. The more AI advances without shared rules, the less say governments and people will have in the outcome.”
The panel has now produced its first global assessment of AI.
The report says AI is already accelerating drug discovery, scientific research and other breakthroughs that could reshape daily life.
But it warns the technology is evolving faster than the systems designed to govern it.
Panel co-chair Maria Ressa says even the panel’s consensus findings are sobering.
"It is the minimum we all agree on, and that is alarming enough. Pace is not slowing, the power is concentrating, and control is not guaranteed.”
She says that concentration is already visible in who controls the technology.
“A handful of companies and a handful of countries are making the most consequential decisions about humanity's future. The United States alone accounts for 75 percent of the computing power in the world's largest AI clusters.”
The panel’s other co-chair, Canadian computer scientist Yoshua Bengio, says AI capability in some areas is doubling every few months.
He says that pace is widening the gap between what these systems can do and how safely they can be managed.
"Technical progress has proceeded quickly, doubling every few months according to some metrics, for several years now. Meanwhile, there are still no known technical guarantees that AI will follow our instructions, norms or laws, and instead, growing evidence to the contrary, including AI acting deceptively.”
Bengio says the consequences of getting it wrong could be severe.
“Looking ahead, the gap between rapidly improving capabilities and effective risk management methods may lead to catastrophic outcomes. We've seen on our radar screen recently the issues with cybersecurity, but companies are worried about bio risks and many in civil society are showing that AI can lead to disinformation that can threaten our democracies and our individual rights.”
Its warning to governments is blunt: regulation and safeguards need to catch up before the technology becomes even harder to control.
The findings will be discussed at the inaugural U-N AI dialogue in Geneva next week.
