AI & human rights — what the Joint Committee’s report could mean for UK businesses

A cross‑party committee has concluded that the UK’s current laws don’t yet protect people from the risks posed by AI. On 14 September 2026, the Joint Committee on Human Rights published a report warning of gaps in the current law and proposing reforms aimed at reducing those risks.

The Government is expected to reply to by 14 November 2026. Although the recommendations aren’t yet law, they indicate how future UK AI regulation might develop.

Here, Sara Ludlam, Eleanore Beard and Megan O’Connor explore the Committee’s key concerns, reforms on the table and steps that businesses should consider now.

Why is the Committee concerned?

While the report recognises that AI can be beneficial, it simultaneously warns of risks that it poses to human rights, including: 

1. Right to equality and non-discrimination — AI is trained and learns from existing data that may underrepresent or contain bias against marginalised groups, such as women and minorities. AI can’t identify bias in the data that it’s trained on, so it treats these biases as fact and reproduces them in its outputs. This raises concerns about discrimination wherever AI is used as its outputs may reinforce existing inequalities. 
2. Right to privacy and data protection — AI relies on large amounts of personal data to produce its outputs, which can create privacy risks. Where AI combines many different datasets to identify individuals from information that was thought to be anonymous but isn’t, there will be an infringement of the right to privacy.
3. Right to an effective remedy — people may not know when AI influenced a decision affecting them. For instance, the report gives an example of AI being utilised to monitor a workplace, resulting in unsubstantiated disciplinary action. This makes it harder to challenge the decision and secure a remedy, therefore undermining the right to an effective remedy. 

What is the Committee proposing?

The report has proposed several legal and regulatory recommendations intended to mitigate the risks identified:

1. A dedicated AI Act

The Committee recommends a dedicated AI Act. This doesn’t exist at the moment, which means that AI is governed indirectly through existing laws not designed to address its specific risks. The Committee believes that a specific AI Act would help to address AI-related human rights challenges directly and consequently strengthen protections.

However, a dedicated AI Act will take time to draft and implement. This means an Act alone can’t address the immediate human rights risks identified by the Committee. Additionally, those in the technology sector worry that excessive regulation could stifle UK tech innovation. This is because AI is never released into the market in its final version, as it receives regular updates and continues to evolve through data training. Any AI Act should therefore consider the nature of the software when considering regulations. 

2. A statutory AI body

The Committee recommends creating a statutory AI body to oversee the use of AI across all sectors. This body would have authority to test new systems before they become available on the market and to withdraw risky systems. 

While a statutory AI body may help to close gaps in current law, having a single regulator covering every sector is ambitious and risks embedding a one-size-fits-all approach. A central body could lack the specialist knowledge needed to understand how AI is used across different industries, as this can vary considerably between sectors such as healthcare and legal services. Having one body could potentially harm the benefits AI brings to particular sectors, which may be heavily regulated already. Therefore, to be effective, the statutory body would need to work closely with existing regulators and establish sector-specific AI approaches.

3. Developers assuming more accountability

The Committee calls for designers and developers of AI systems to have more accountability. They believe liability currently disproportionately falls upon businesses implementing AI into their practice, rather than the companies building the AI systems. The companies creating the AI systems are most capable of preventing infringements on human rights because they’re designing the software. Placing more accountability on the designers would incentivise them to produce the systems in a way that minimises risks to human rights. 

While greater accountability for developers is welcome, businesses must remain responsible for how they use AI. Through accepting an AI supplier’s terms and conditions, businesses are presumed to understand a system’s risks and therefore accept accountability for using its outputs. 

As Goldman Sachs CEO David Solomon notes, accountability for automated decisions has been established in finance for decades. He describes automated decision-making as well established, noting that Goldman Sachs has used it on its trading platforms for 20 years, with “human accountability attached to every one of those platforms”. Accountability for automated decisions is therefore not entirely new territory and those approving an automated decision should take responsibility for approving its outputs. Accountability should be shared between developers and users, as this would encourage the strongest protection of human rights.

4. Regulations based upon risk

The Committee argue that the number of regulations imposed on an AI system should be dependent upon the level of risk that an AI system poses. This means that riskier systems would have greater regulations placed on them and the most harmful systems would be banned altogether. The level of risk would be assessed by considering the amount of harm that they pose to human rights. The Committee said, “strong sanctions [should be imposed] for those companies and organisations who damage human rights through their use of AI systems”. 

5. Transparency & explainability

The Committee want organisations to be more transparent when AI is being used to make a decision about them (for example, when AI is influencing recruitment). At the moment, people don’t need to be made aware when AI is used in a decision-making process. The committee want to change this, stating that organisations should be more transparent when decisions are assisted by AI. and those decisions should be accompanied with sufficient information explaining why the AI system came to that conclusion. This will enable people to more effectively challenge those decisions if they’re deemed incorrect or unfair, supporting the right to effective remedy. Businesses can be more transparent through introducing AI usage policies, recording staff AI training and documenting AI risk assessments.

6. Council of Europe Framework

The Committee urges the Government to publish a clear timetable for ratifying the Council of Europe Framework Convention on AI into UK law. This is a broad international treaty that the UK government has signed but not ratified, which aims to protect human rights, democracy and the rule of law as AI continues to be widely adopted. 

To do this, seven principles must be enforced:

1. Human dignity and individual autonomy.
2. Transparency and oversight.
3. Accountability and responsibility for harmful or discriminatory AI outcomes.
4. Equality and non-discrimination.
5. Privacy and data protection.
6. Reliability and safety.
7. Safe innovation.
    

The Convention details remedies those harmed by AI can access and requires governments to manage risks AI poses to human rights. The Committee believe that ratifying this framework would commit the UK into policies that promote AI systems to respect human rights.

Will tougher rules hinder innovation?

The Committee believes that “none of this need compromise the use of AI to help generate economic growth”. However, technological innovation drives economic growth and excessive regulation could push developers towards jurisdictions with fewer restrictions. While AI does require careful supervision, overly restrictive rules could undermine its economic and operational benefits. 

TechRound — an online technology magazine for businesses and start-ups — highlighted concerns that stronger AI regulation could place a disproportionate burden on smaller companies. Patrick Gibbs of Epiphany Dynamics warned that they could face “compliance costs designed for a completely different category of company” and argued that a bill should distinguish between smaller and larger AI operators.

Lord Clemente-Jones (Liberal Democrat peer and spokesman for the digital economy in the House of Lords) posted on LinkedIn that he “warmly welcome[s]” the report and that “Growth and human rights protection are not in tension — they are, done properly, the same project”. He says that he’ll be pressing to put these recommendations into practice in the months ahead.

What would the Committee’s recommendations mean for businesses if implemented?

If the Committee's recommendations are implemented, there would be a big increase in businesses loss, risk and liability. Businesses would need to identify and closely monitor the extent of AI use in their organisations. To do this, they should maintain human oversight, keep evidence and records of its use and consider safeguards that they could implement against potential discrimination.

Why businesses should do now

Businesses that proactively prepare for possible future regulations will be in a stronger position than businesses that simply wait for regulations to be enacted. 

What to do next:

• Look out for the Government’s response to this report (estimated by 14 November).
• Read the report for more information.

Talk to us

If you want to understand how your business can use AI responsibly while protecting human rights, our tech sector and data protection experts can help you to assess AI-related risks, strengthen your policies and safeguards and prepare for regulatory developments. 

Talk to our team by giving us a call on 0333 004 4488, sending us an email at hello@brabners.com or completing our contact form.