How to close AI’s accountability loophole - FT中文网
登录×
电子邮件/用户名
密码
记住我
请输入邮箱和密码进行绑定操作:
请输入手机号码,通过短信验证(目前仅支持中国大陆地区的手机号):
请您阅读我们的用户注册协议隐私权保护政策,点击下方按钮即视为您接受。
人工智能

How to close AI’s accountability loophole

Governance of new technologies must be determined by elected officials rather than fastest moving companies
00:00

{"text":[[{"start":6.55,"text":"The writer was director of the national AI office under President Joe Biden and US Ambassador to the OECD from 2009-12"}],[{"start":15.100000000000001,"text":"Last Thursday, the White House scrapped an executive order that would have standardised pre-release safety testing of advanced AI models. Once again, those advocating a trust-but-verify approach lost to the no-holds-barred camp."}],[{"start":29.05,"text":"AI currently operates in an accountability-free zone. As it replaces existing systems throughout the economy, its black-box algorithms cannot easily be bound by rules written for humans. The question is no longer whether these tools will proliferate, but whether the terms are set by democratic choice or the companies moving fastest. The risks of getting this wrong are serious. Chatbots giving medical advice face neither the malpractice liability nor the licensing oversight of human doctors. In the US, governments can use error-prone AI systems to deny benefits without citizens’ recourse to explanation from or appeal to a human."}],[{"start":70.25,"text":"US courts are struggling to apply existing law. A lawsuit against OpenAI alleges product liability and competition violations, which it denies, for conduct for which “every therapist, teacher, and human being would face criminal prosecution.” The court must decide whether a chatbot is even a product, and if it can be held liable for competing unfairly with humans by doing what humans legally cannot."}],[{"start":95.9,"text":"This “accountability arbitrage” to gain market dominance is not new. US social media companies perfected the playbook, using a legal loophole to sidestep media norms that require news outlets to carry liability insurance and uphold editorial standards. Uber said it was a technology platform, not a taxi company; its drivers were gig workers, not employees. Airbnb argued it owed no hotel taxes and was not bound by zoning or building-safety rules."}],[{"start":125.55000000000001,"text":"As AI spreads, it threatens to open a society-wide loophole in democratic governance. The laws around human conduct hold people to account for the harms they cause. If algorithms sit outside those rules, and no new standards take their place, the choice of which harms to tolerate falls, by default, to the first movers. Without public standards, competitive pressure rewards corner-cutting. In February, Anthropic softened its policy to guarantee safety measures before release, noting that without rules for all those “with the weakest protections would set the pace”. "}],[{"start":158.9,"text":"Past US governments built new frameworks for new technologies. The Nuclear Regulatory Commission approves construction, inspects plants and demands incident reporting; operators face strict, no-fault liability. Companies making cars, drugs and aircraft are governed by a mix of existing tools, including liability, and purpose-built regimes that are imperfect, often slow and frequently captured. Such flaws are an argument to modernise them, alongside new AI-specific safety rules, not to abdicate public governance."}],[{"start":193.05,"text":"After ChatGPT’s launch in late 2022, several governments took first steps. President Joe Biden signed an executive order in October 2023 directing federal agencies to address AI in their sectors and to test the most advanced models. Days later, at Bletchley Park, 28 countries, including the UK, US, China and EU member states, agreed to co-operate on frontier-AI risks. The EU passed its AI Act in May 2024."}],[{"start":222.8,"text":"The momentum in the US has shifted. Donald Trump, elected with industry support, revoked Biden’s order and has focused on bolstering domestic AI giants to stay ahead of China. The order scrapped last week would have slightly softened that stance. Even that was too much."}],[{"start":239,"text":"Progress elsewhere faces stiff resistance. After the US Department of Justice joined a lawsuit by xAI, Colorado watered down its first-in-the-nation law that would have imposed pre-deployment responsibility on AI systems involved in consequential decisions such as those concerning employment or housing. The venture-capital firm Andreessen Horowitz and industry allies poured more than $100mn into political spending against state AI rules, including over $3mn to defeat the congressional candidate who co-wrote New York’s safety-plan law. "}],[{"start":273,"text":"Polls indicate growing concern about unfettered AI. On Monday, Pope Leo XIV endorsed a central role for governments, writing, “it is necessary to establish adequate regulatory tools capable of upholding justice and curbing the distorting effects of technological power.” He rightly accuses most people of, “watching and waiting . . . merely hoping for the best”. Instead, citizens and elected representatives have a clear responsibility to update governance mechanisms so that AI rewards innovation, rather than evading accountability."}],[{"start":310.55,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1780054162_6867.mp3"}

版权声明:本文版权归FT中文网所有,未经允许任何单位或个人不得转载,复制或以任何其他方式使用本文全部或部分,侵权必究。

太空探索技术公司IPO大胆踏入AI经济学的未知领域

对于前沿模型最基本的商业模式,我们仍有许多有待了解之处。

教宗搅动硅谷

不同于美国总统,这位教宗选择直面AI带来的严峻挑战。

科学家发现:鸟类凭“直觉”导航

研究发现,肝脏中富含铁的免疫细胞像一个“内置罗盘”,帮助鸽子感知地球磁场。

法拉奇正面临来自右翼的威胁

“复兴英国”的崛起或将迫使改革党领导人调整策略。

如果问题根源在远程办公而非AI,这对初级岗位招聘意味着什么?

最新证据显示,居家办公的兴起让企业对聘用初级员工的意愿下降。

被罢免的英国石油董事长阿尔伯特•马尼福德在遭撤职前曾与公司秘书发生冲突

在这家石油巨头发生剧变后,本•马修斯已请假离岗。
设置字号×
最小
较小
默认
较大
最大
分享×