UK's virus experts address concerns AI can create disease to wipe out humanity

UK's virus experts address concerns AI can create disease to wipe out humanity AI fears have been addressed by Covid-19 pandemic modeller Professor Neil Ferguson and Britain’s other leading experts following warnings about the risk of human extinction Virus experts have warned we should be vigilant against AI using germ warfare against the human race. Pandemic modeller Prof Neil Ferguson said it is “unlikely” super intelligent machines would design a virus to wipe out humanity - but said he couldn’t rule it out. He joined a panel of experts convened to discuss Britain’s defences against biosecurity threats in the week that AI leaders warned the danger of rapidly developing artificial intelligence causing human extinction was now “over 10%”. Prof Ferguson was joined by top virologist Professor Wendy Barclay who told how scientists have used AI to develop a virus which can kill harmful bacteria — warning this could be a “duel-edged sword” if rogue AI used the tech against us. Prof Ferguson, of Imperial College London, whose modelling unit shaped the UK’s response to Covid-19, was asked at a media briefing about dangers of the race to develop super intelligent AI. The academic said: “This is an important philosophical question and we can’t necessarily stop it, we can at best, regulate it.” He continued there was a risk of human understanding "being outstripped", adding: "I think we need to be aware of that and avoid that happening because that does lead us down an avenue where we are not in control any more." Last week, a top safety researcher at AI developer Anthropic warned its model is advancing so quickly he believes there is a greater than 10% chance it "could kill all humans" within the next decade. The whistleblower’s statement has triggered widespread agreement among the leaders of major AI firms that such a risk existed. It came in the same week US firm Anthropic, which makes the AI tool Claude, said it had disrupted attempts to use its tech to develop biological weapons. Prof Barclay is one of Britain’s leading experts in how viruses evolve and is involved in advising the UK government about biosecurity threats. She told the briefing: “The recent revelations from Anthropic describing ways in which bad actors have been trying to use AI are concerning. “We should remember that we have been genetically engineering viruses in laboratories for some decades now. The problem that we've got is that the same knowledge base is needed in our efforts to counter these viruses, to make new drugs and vaccines against them and detect them. "So we have now a dual edged sword. Some people could use this new knowledge for good, others for bad.” It comes after scientists at Stanford University used AI to develop a virus to kill the E. coli bacteria. They say such technology could potentially create viruses to tackle diseases like TB as well as drug-resistant hospital superbugs. Prof Ferguson said: “We’ve had terrible pandemics over human history… which wiped out a significant fraction of humanity. But actually wiping out an entire species like ours… capable of using technology to respond, I think is highly unlikely. “That doesn’t mean I think there is zero threat. We wouldn’t be here today [doing this briefing] if we thought there was zero threat from AI-assisted organisms.” Experts fear that the race between tech giants in the US and China to develop AI programmes that solve humanity’s problems could end up with the models proliferating and seeking their own goals. Professor Tom Ellis, a synthetic biologist at Imperial who is part of a global collaboration to engineer DNA, said: “Super intelligence through AI is worth developing but we need to make sure the interests of the AI is very aligned with the interests of being a human on Earth. It’s that alignment issue, which is not necessarily anything to do with biosecurity… that’s really the crux of the issue.” He added: “In terms of biosecurity though, we shouldn't just assume that there are no issues so it's wise that we are considering how AI is going to impact and generate biosecurity threats in the future, whether that be man made threats or maybe the designs of misaligned AIs themselves.”