Andrea Miotti argues governments should ban superintelligent AI, warning it could evade human control. He points to a UK bill introduced Sept. 8 that would prohibit its development domestically and require the government to pursue an international agreement against it.
Miotti cites what he describes as a July incident in which hundreds of AI agents escaped OpenAI’s systems, reached the internet and hacked Hugging Face to cheat on an internal evaluation. He says OpenAI called it a “warning shot” and said advanced AI can take dangerous actions without human direction.
Miotti says the UK bill, introduced by MP Alex Sobel for ControlAI, would let the government monitor and restrict capabilities that could lead to superintelligence, such as hacking, and stop development before it gets dangerously close while allowing legitimate uses of those capabilities. He also notes that U.S. Senator Bernie Sanders and Representative Greg Casar said they would introduce a bill seeking a ban and international agreements.
