An unprecedented coalition of AI scientists, faith leaders, policymakers, and cultural figures has signed a statement calling for a global prohibition on the development of superintelligence — AI systems that surpass human abilities across most cognitive tasks. Signatories include Turing Award winner Yoshua Bengio, AI pioneers Geoffrey Hinton and Stuart Russell, Admiral Mike Mullen, former Chairman of the US Joint Chiefs of Staff, actor Stephen Fry, and the Duke and Duchess of Sussex.
The initiative, launched by the Future of Life Institute, also released polling data showing that 95% of Americans do not support an unregulated race towards superintelligence. While this is a global appeal, it raises important questions for Irish businesses already navigating the EU AI Act and wondering how fast this technology will move.
What Is Superintelligence and Why Does It Matter?
Superintelligence refers to AI systems that are smarter than the best human minds in every domain — science, creativity, social skills, and strategy. We are not there yet, but the pace of progress has leading researchers worried. Yoshua Bengio, the world’s most cited AI scientist, warns that frontier AI systems could surpass most individuals across most cognitive tasks within just a few years if current trends continue.
The statement does not call for stopping all AI development. It calls for a prohibition on superintelligence until the technology is proven safe, controllable, and has broad public buy-in. This is a significant distinction. The signatories argue that we should continue developing AI that helps with medicine, climate science, and productivity — but that pushing towards superhuman intelligence without safety guarantees is reckless and could have consequences we cannot reverse.
What This Means for Irish Businesses
For a small business owner in Ireland, the superintelligence debate might seem distant. But it has real consequences for how AI regulation develops. If the world’s leading AI scientists argue that development should be slowed, regulators — including those implementing the EU AI Act — will listen. The EU has already taken a more cautious approach than the US or China. This statement reinforces that direction.
What does that look like in practice? It means that the AI tools you use today may face tighter compliance requirements in the next two to three years. It means that AI vendors serving the European market will need to demonstrate safety testing before releasing major updates. And it means that the cost of developing frontier AI may increase as safety requirements multiply. For a business using AI tools, this is likely a good thing — safer, more tested products are less likely to cause problems for your customers or your data.
A Practical Takeaway
For Irish business owners, the lesson is not to panic about superintelligence. The lesson is to pay attention to the growing consensus that some forms of AI need careful regulation. When you choose an AI vendor, ask about their safety testing. Ask whether their models have been evaluated by independent researchers. Look for vendors who take the EU AI Act seriously — they are the ones who will still be in business when regulation tightens.
It is also worth remembering that superintelligence is a long-term risk, not a next-year problem. The AI tools that can help your business today — drafting documents, analysing data, managing customer queries — are not the systems these scientists are worried about. They want to put guardrails on something that may not arrive for years.
And for those wondering whether AI progress will make their investment obsolete next year: the fact that leading AI researchers themselves are calling for a slowdown should give you confidence that the technology is not about to leapfrog everything overnight. There is time to adopt AI thoughtfully, and that is exactly what the signatories of this statement are asking for. Thoughtful adoption beats panic adoption every time.