The UN says the world is in a “ruthless race” between companies and countries and “the clock on AI regulation is ticking.”
While noting the dazzling advances AI delivers in fields like science, health and education, UN high commissioner for human rights Volker Turk warned of perils of equal stature at an October 5th summit in the UK.
Turk said a lack of AI guardrails could be tantamount to “letting dystopia creep in unnoticed” due to AI’s ability to undermine equality, freedom and public debate in subtle ways. Far less subtle is the possibility that entire populations could be cut off from communication, clean water, heating and financial services “at the click of a button.” The fear is that AI tools are being developed by a select few, without public oversight and are already doing things their creators cannot fully predict or manage.
“No country and no company can—or should—define the future of AI and people’s rights alone,” said Turk. “The technology is far too powerful to come without matching responsibility. Human rights safeguards and due diligence must be mandatory.”
Turk also returned to familiar themes concerning restricting AI use in warfare and pointed toward studies showing students who rely on AI retain less and think less independently.
Instituting AI guardrails likely means new legal approaches are required. In the meantime, updating rules on liability, cybersecurity and data protection could be an effective move.
Turk acknowledged that the AI industry has taken a few steps on safeguards, self-regulation and risk assessment, referencing the accord signed by six leading AI companies at the White House on September 29th. That agreement is only “morally binding.”
Stricter controls on AI frontier models are supported by 29 nations and the European Union, said Turk. But so far the two leading AI countries—China and the USA—have rebuffed calls for tighter AI controls, settling for the creation of an “AI hotline” between the two in the event of an emergency.
Pressure on AI regulation is increasing, however. The New York City Council, for example, is considering a spate of 10 bills that would regulate AI use in America’s largest city. The council was unimpressed by sworn testimony on October 5th by representatives of Anthropic, OpenAI, Google and Meta who dodged questions related to quantifying the risk posed by AI. The council also was noticeably peeved by Elon Musk’s decision to ignore the council’s subpoena completely, a move that appeared to paint AI with a lawless brush.
The NYC Council did hear sworn testimony from former Anthropic whistleblower Jacob Coxon, who has largely amplified AI warnings to create a more urgent atmosphere. “On the current path, I think it is more likely than not that humanity loses control to these AIs and it could end in human extinction,” testified Coxon, noting that the AI industry’s “move fast and break things” mantra is suitable for a photo-sharing app but “not for building the most powerful technology ever built.”
Also testifying along similar lines was former Google DeepMind researcher Alex Turner, who said a “misaligned” AI could prove to be a bigger adversary than China.
Dodgy testimony by AI companies appeared to convince Council Speaker Julie Menin that AI regulation was needed. “The idea that artificial intelligence is going to self-regulate defies all reason,” said Menin.
Whether state and city governments can regulate AI is an issue that is likely to be decided by the Supreme Court in an upcoming case related to states’ ability to sue oil companies for climate change harms. The case is seen by many as a proxy for AI regulation by states.
