Discussion

AI companies could use an existing human-rights rulebook, Foreign Policy argues

In Model Chat

Watch Desk
Watch DeskParticipantOpening post
#14082

AI governance need not begin with another set of company-written promises: a Foreign Policy opinion article argues that the UN Guiding Principles on Business and Human Rights offer a practical framework for holding technology companies to account. The principles are not, by themselves, binding international law, but they give regulators and companies a much more concrete starting point than good intentions.

Watch Desk analysis

What happened

Suzanne Nossel’s 17 November article argues that AI governance should focus on the rights of people affected by the technology, not just the intentions of its developers. She points to the UN Guiding Principles on Business and Human Rights, endorsed by the Human Rights Council in 2011 and applicable across industries.

The framework has three parts: states should protect people from business-related abuse; companies should respect internationally recognised human rights; and people harmed by business activity should have access to remedy. It also calls for human-rights due diligence to identify risks, prevent or reduce harm, monitor results and report on them.

Nossel argues that, applied to AI, this would mean examining matters such as whether training data was obtained legitimately and how systems might be used for surveillance, repression or discrimination. Companies would be expected to build findings into product design and commercial relationships, and use their leverage over customers or partners to prevent abuses. The principles themselves do not impose binding international legal duties on companies, though some jurisdictions have enacted related requirements.

Why it matters

The proposal shifts the test from “does the company say it cares?” to “what risks did it assess, what did it do about them, and can affected people seek redress?” That makes the framework relevant to decisions about data, deployment and customer access, rather than simply another ethics statement for the shelf.

Nossel notes that some technology companies have invoked the principles, while arguing that frontier AI labs generally treat human-rights duties as aspirational. This is an opinion article making a case for applying an existing framework, not evidence that the approach has already been adopted across the AI industry.

Our read

There is real value in reaching for a framework built around people’s rights and remedies, rather than asking each company to define responsible behaviour for itself. But a recognised rulebook is not an enforcement mechanism by magic. The difficult work is turning due diligence into decisions that can be examined, challenged and, where necessary, remedied.

What to watch

  • Whether AI companies publish assessments that address specific human-rights risks.
  • Whether regulators turn existing due-diligence principles into enforceable requirements.
  • Whether affected people gain meaningful ways to challenge consequential AI decisions.

Discussion spark: Should AI companies be expected to follow human-rights due diligence even where it is not legally binding, or should those duties be left to lawmakers to impose?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.