David Lammy says the UK needs to harmonize rules with other countries to govern the use of AI in nuclear deployment decisions. Christopher Furlong/Getty Images
The Labor government may strike a pact with other countries to jointly ban the use of artificial intelligence (AI) in nuclear weapons deployment decisions, reports The Telegraph.
In a speech on Monday, David Lammy, the shadow foreign secretary, will say that the UK needs to harmonize rules with other countries, including China, to regulate the use of AI in relation to nuclear weapons.
The rapid development of AI in the past year has left governments scrambling to figure out how to safely control the technology.
Concerns have been raised about whether AI could be introduced into the decision-making process about when to deploy nuclear weapons.
Countries may choose to automate elements of their nuclear command and control to increase deterrence, ensuring that retaliation is sure to strike if they are under nuclear attack.
Last week Dame Barbara Woodward, British Ambassador to the United Nations , highlighted this as a concern, announcing that the UN Security Council will discuss AI for the first time later this month.
She said: “There should be concerns about how AI could be used in, for example, nuclear technology – whether any country with nuclear weapons would consider outsourcing control of those weapons to AI and how that could go wrong – and that the same with automatic weapons.
Isn't artificial intelligence developing too fast? Poll
In his speech at the Bingham Center for International Law on Monday, Mr. Lammy will indicate that the Labor government will seek to reach agreements with other countries on this issue.
He is expected to say: «Artificial intelligence offers incredible growth opportunities, but this potential will be undermined if governments do not act together to prevent the highest risks it poses.»
«Besides domestic regulation , we must lead a global diplomatic effort to prevent the use of AI by terrorists, criminals and autocrats who hope to use it to suppress freedom, spread disinformation and undermine democratic processes.
“We must work with the widest possible coalition of countries to develop and manage AI, including in the areas of technology impact on nuclear weapons.
«It's in Britain's best interest to include China, because it makes no sense to exclude one of the largest AI markets from these rules.»
«Probable paths for the destruction of humanity»
Lord Haig, former Foreign Secretary, has previously stated that it is entirely possible to reach an agreement with China not to hand over control of nuclear weapons to AI.
“There are things that countries need to agree on, such as that AI will never be allowed access to nuclear weapons or… dangerous pathogens,” he said earlier this year. “It is ethically and politically necessary, right now, that you never allow this to come near these things, because these are the most likely ways to destroy all of humanity.”
Governments are already taking the first preliminary steps. steps towards international regulation of AI. Last month, The Telegraph reported that Rishi Sunak would be hosting a global summit in London later this year aimed at developing international rules for AI.
This follows a warning from the heads of the world's most advanced AI lab, OpenAI. Google Deepmind and Anthropic — that «reducing the risk of AI extinction should be a global priority along with other societal risks such as pandemics and nuclear war.»
At a Downing Street meeting with the Three Bosses in May, Mr. Sunak discussed the risk of technology, «ranging from disinformation and national security to existential threats.»
Свежие комментарии