Should governments restrict AI development over fears of superintelligence?
Leading AI scientists including Geoffrey Hinton and Yoshua Bengio, along with some company executives, warn that superintelligent AI could pose existential risks to humanity through loss of control or misuse. In September 2026 US lawmakers Bernie Sanders and Greg Casar proposed the Ban Artificial Superintelligence Act to prohibit systems exceeding human intelligence and pause advanced AI pending safety rules. The UK government is instead investing over £1 billion in AI hardware, compute and adoption to boost growth while developing regulatory sandboxes.
19 Opinions
This for me was difficult, AI was the potential to be extremely useful, but there has to be regulations in place to ensure it cannot be used for illegal purposes, it must not be able to affect lives in a negative way, jobs being lost. People must always be in control in a sensible way. Until such time I said yes to this.
Much like Net Zero, if we take a puritan view on AI we risk falling greatly behind in the chance to create jobs, industry and increase GDP. Superintelligence is a risk, but Humans kill enough Humans daily too.
I don't think government could restrict it if it tried. Better to manage risks another way
YES ...but just temporarily. Not necessarily restrict, I'd say pause until we're sure that there's checks & balanced in place to stop the robots taking over. As we've already seen cases where the nightmare/move scenario has almost happened, admitted not to the same extent, but who knows in the future!
Evidence suggests our current government ( and the ones before it ) couldn’t organise themselves out of a paper bag. Perhaps we should listen to the scientists ( as long as they are properly qualified and not the modelling ‘experts’ often employed) and approach this area with caution. I also feel that excessive water usage is a danger - considering the mess our water companies seem to be in.
Yes, but with heavy caveats. Governments, in general, have proven themselves woefully inadequate at dealing with the rapid spread of new technology. AI is the latest, and largest, issue that we are ill-equipped to handle. A British AI council, comprised of elected officials and industry experts, should oversee any legislation aimed to control development in this country. There are myriad risks with AI, but we do need to keep pace with the rest of the world.
AI development should be restricted, but by aims ie we should focus on certain limited applied practical uses rather than saying " I did a thing, but I don't know what to use it for- faster tiktoks?".
The geeks have already let the genie out of the bottle before the bureaucracy has even reached for its shoes. The tech bros have financially too much at stake to hold back and the financial establishment lent far too much to stop. We are now playing catch-up, a game society looks likely to lose.
AI development should be restricted in this country, but fears of superintelligence would be a long way down my list of why. Energy use, water use, and the deleterious effect AI development is having on the market for computing power used for literally any other function are much bigger reasons. We are burning our resources at an alarming rate all so that we can generate misinformation quicker than ever before.
The genie is out of the lamp at this point. All the necessary white papers are public. Bad actors are already using AI to carry out sophisticated hacks with swarms of autonomous processes. Someone stole $100M of Bitcoin like this: Google "Coldcard hack". Ask Google Gemini for a simple explanation of the "Hugging Face hack". That was an escape from OpenAI, a frontier lab that uses guardrails but made a mistake. The only defence is "Defensive Co-scaling": AI vs AI