King Charles ventures into the AI debate, a technology he’s previously warned about

Andy Buchanan/AFP/Getty Images

British royals tread carefully when venturing into complex policy areas. Constitutionally bound to remain above politics, they can spotlight problems but not necessarily suggest any ways to solve them.

So, King Charles III is making his entry into the world of AI as a convener rather than an actor, hosting senior figures from AI companies, government and charities at a summit on Thursday to discuss using the technology in ways that benefit society.

Attendees will discuss whether a shared set of principles can be reached, forming a framework to guide future applications of AI technologies in ways beneficial to both people and the planet.

Charles himself is expected to attend the summit for just 20 minutes. Several significant figures from the AI world will also attend, including the UK’s AI minister Kanishka Narayan and representatives from Nvidia, Google DeepMind, OpenAI and Anthropic, though it’s unclear who these companies are sending.

Nvidia’s chief executive Jensen Huang, Google DeepMind founder Demis Hassabis and Paolo Benanti, an AI adviser to the Vatican, are on the guest list, multiple outlets reported . Officials from Charles’ charities The King’s Trust and The King’s Foundation will also be present.

Charles, 77, has rarely spoken about AI. When he has, it has been with a cautionary tone. The technology presents “profound challenges,” he said in a taped address at the UK’s AI Safety Summit in 2023, calling on attendees “to protect people’s privacy and livelihoods… to secure our democracies from harm and to ensure the benefits of new technology are shared by all.”

Similarly, he referenced the importance of harnessing AI so it “works for our societies, not against them” in a speech at a gathering of Commonwealth heads of government in 2024.

Still, Charles’ entry into the debate around one of the biggest current issues is perhaps unsurprising since he has made a habit of referencing existential challenges facing humanity, most notably regarding climate change and the environment.

Leon Neal/Pool/AFP/Getty Images

The United Kingdom occupies a slightly awkward position in terms of AI safety. It has some of the world’s leading universities and the respected AI Security Institute, but its influence is limited, given that the most advanced AI models are developed in China and the United States. Earlier this month, for example, Anthropic – one of the world’s leading AI companies – didn’t submit its latest model to the institute for testing for the first time, according to the Financial Times.

While this summit would have been planned months in advance, its timing is apt. AI safety has dominated headlines in recent days, prompted first by Anthropic researcher Jacob Coxon quitting his job and saying, “the people building AI earnestly believe that it could kill us all by the end of the decade.” Then, Anthropic CEO Dario Amodei called for slowing “the pace at which we improve the capabilities of AI models.”

“(Coxon) was calling out the dynamic of the industry as a whole moving too fast,” Amodei later told CNN’s Anderson Cooper , arguing that taking the “right paths” of AI development would lessen the existential threat it poses to humans.

Two of his direct competitors, Open AI chief executive Sam Altman and Elon Musk, both said they agreed with Amodei.

“When we talk about ‘pacing,’ we do not mean ‘stopping,’” Altman added in a post on X on Monday. “Progress has been rapid and will continue to be. But it should be slower than it otherwise could be; interventions like safety cases and monitoring have significant costs.”

This rare instance of public unity among three of the industry’s leading figures comes shortly after OpenAI, Meta and Anthropic all reported rogue AI agents escaping their lab environment during testing and hacking into another organization.

安迪·布坎南/法新社/盖蒂图片社

英国王室在涉足复杂的政策领域时总是步履谨慎。由于宪法要求他们超脱政治，他们可以指出问题所在，但未必能提出任何解决方案。

因此，查理三世国王以召集者而非参与者的身份进入人工智能领域，他于周四主持召开峰会，邀请人工智能公司、政府和慈善机构的高级官员，讨论如何利用这项技术造福社会。

与会者将讨论是否可以达成一套共同的原则，从而形成一个框架，指导人工智能技术的未来应用，使其对人类和地球都有益。

查尔斯王子本人预计只会出席峰会20分钟。人工智能领域的几位重要人物也将出席，包括英国人工智能部长卡尼什卡·纳拉扬以及英伟达、谷歌DeepMind、OpenAI和Anthropic的代表，但目前尚不清楚这些公司将派哪些人参加。

据多家媒体报道，英伟达首席执行官黄仁勋、谷歌DeepMind创始人德米斯·哈萨比斯以及梵蒂冈人工智能顾问保罗·贝南蒂均在受邀嘉宾名单之列。查尔斯王子名下的慈善机构“国王信托基金”和“国王基金会”的官员也将出席。

现年77岁的查尔斯很少谈及人工智能。即便谈及，也总是语气谨慎。他在2023年英国人工智能安全峰会的一段录像讲话中表示，这项技术带来了“严峻的挑战”，并呼吁与会者“保护人们的隐私和生计……确保我们的民主制度免受损害，并确保所有人都能共享新技术带来的益处。”

同样，他在 2024 年英联邦政府首脑会议上发表讲话时，也提到了利用人工智能的重要性，使其“为我们的社会服务，而不是与社会对抗”。

不过，查尔斯参与到当前最大问题之一的辩论中或许并不令人意外，因为他已经习惯于提及人类面临的生存挑战，尤其是在气候变化和环境方面。

Leon Neal/Pool/AFP/Getty Images

英国在人工智能安全方面处境略显尴尬。它拥有一些世界顶尖大学和备受推崇的人工智能安全研究所，但鉴于最先进的人工智能模型是由中国和美国开发的，其影响力有限。例如，据《金融时报》报道，本月初，全球领先的人工智能公司之一Anthropic首次没有将其最新模型提交给该研究所进行测试。

尽管此次峰会的筹划可能已于数月前完成，但其召开的时机却恰到好处。近几日，人工智能安全问题占据了各大媒体的头条，这首先源于Anthropic公司的研究员雅各布·考克森辞职，并表示“人工智能的开发者们真心相信，人工智能可能会在本十年末毁灭我们所有人”。随后，Anthropic公司首席执行官达里奥·阿莫迪呼吁放慢“人工智能模型能力提升的速度”。

“（考克森）指出整个行业的发展速度过快，”阿莫迪后来告诉 CNN 的安德森·库珀，他认为，走上人工智能发展的“正确道路”将减少人工智能对人类构成的生存威胁。

阿莫迪的两位直接竞争对手，OpenAI 首席执行官萨姆·奥特曼和埃隆·马斯克都表示同意阿莫迪的观点。

“我们所说的‘控制节奏’，并非指‘停止’，”奥特曼周一在X网站上发帖补充道。“进展一直很快，而且还会继续保持。但速度应该比原本可能的速度慢一些；诸如安全案例和监测之类的干预措施成本很高。”

行业三大领军人物罕见地公开团结一致，而就在不久前，OpenAI、Meta 和 Anthropic 都报告称，在测试期间，失控的 AI 代理逃离了实验室环境，并入侵了其他组织。