UK calls for some AI uses to be banned outright
Key Points
- Joint Committee on Human Rights wants some AI uses prohibited outright
- Subliminal techniques and misuse of profiling or biometric data named as candidates
- Public consultation proposed on the detail of what to ban
- High-risk AI systems would need prior approval before deployment
- A human in the loop is not enough to count as meaningful oversight, the report says
The Joint Committee on Human Rights has told the government to ban some uses of AI outright because they are incompatible with human rights.
In a new report on human rights and the regulation of AI, the committee named subliminal techniques and inappropriate use of profiling or biometric data as potential areas for prohibition.
It said the government should hold a public consultation on the detail of which AI uses should be prohibited.
Other AI systems that pose a high risk of causing human rights harms should need prior approval before companies can deploy them, the committee said.
The bans would form part of a proposed new AI Bill that takes a risk-based approach to AI harms. Under the committee’s model, low-risk AI systems would face fewer and less demanding legal requirements, while high-risk systems and models would carry more demanding obligations.
The committee said legislation should not place unjustified burdens on businesses, so that UK companies can innovate and take advantage of the benefits AI can offer.
“We also recognise that AI has the potential to provide great benefits to society, including improving how human rights are protected,” said Alex Sobel, chair of the Joint Committee on Human Rights. “That is why it will be important to ensure that regulation is proportionate and targets the areas of AI development and usage that pose the greatest risks.”
The report said the risks of AI harms are particularly acute for already minoritised communities, including Black and minority ethnic people, and that the right to privacy is under serious threat.
The committee also called for mandatory transparency requirements across the entire AI lifecycle. Any AI system that can have a significant impact on individuals, groups or communities should carry an obligation to state when and how it is being used.
On automated decision making, the committee said UK GDPR rules need strengthening to provide robust protections against misuse. It said the mere presence of a human in the loop is not enough to count as meaningful human involvement or intervention in a decision.
“We need these protections in now, it cannot wait until fear human rights risks become reality,” said Sobel.
