FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Apex Insurance Brokers
Speak to a brokerGet a quote →

Aipedia: Social workers

AI in adult social care eligibility decisions

AI can support adult social care eligibility and resource decisions, but significant decisions about people's care need meaningful human involvement and a route to challenge. The social worker's judgement and recording remain central.

In short

Councils are exploring AI to help sort assessments, suggest eligibility outcomes and estimate personal budgets or support hours. These are significant decisions about people's lives, often based on health and disability data. UK GDPR Articles 22A to 22D, in force from 5 February 2026, restrict solely automated significant decisions on special category data and require safeguards including human intervention and the right to contest. Social Work England's standards require professional judgement and records that show how decisions were reached. Responsibility for the tool sits with the employer, but the social worker who signs off the outcome is still accountable for their part.

How AI is used in eligibility and allocation

Tools in this area range from drafting aids that turn assessment notes into eligibility summaries, through to systems that suggest whether needs meet the threshold or propose a level of support or budget. SWE's research found that use of AI across social work is uneven and that employer policies vary, so practitioners may meet very different tools depending on where they work.

The appeal is consistency and speed. The risk is that a suggested outcome becomes the default and that the reasoning behind it is hidden inside the tool.

The duties that apply

These duties apply whether or not AI is involved:

SWE's executive director, writing on the GOV.UK social work with adults blog, said AI cannot replace professional judgement and warned that biased data can wrongly label people "low priority".

Challenges by service users

People who disagree with an eligibility decision or a reduction in support may complain, ask for a review or seek to challenge the decision formally. Where AI played a part, they may ask what the tool did and whether a person really considered their case. If the record shows only a tool's output and a sign-off, the decision is harder to defend. If it shows the social worker's analysis of the person's own account, the evidence and the outcome, it is much stronger.

What can go wrong

Illustrative scenario: an AI tool suggests reduced support hours following a review, based on structured answers that do not capture a person's fluctuating condition. The social worker accepts the suggestion without testing it against what the person and their carer said. The person's situation deteriorates and their family challenges the decision, asking whether anyone actually considered the fluctuation. The record does not show that anyone did.

Employer versus individual accountability

SWE's research found that accountability between social worker, line manager and software maker is often unclear, and that employers should procure responsibly, set policies, quality-assure outputs and supervise. Legal responsibility for the council's decisions and its tools sits with the council, and claims are usually handled through its arrangements. The individual social worker remains accountable to SWE for their own judgement and records; see AI errors and fitness to practise.

Independent social workers carrying out reviews or assessments for councils should check whether their contract makes them responsible for outputs from the council's tools, and how their PI treats that work; see independent social workers and AI and does PI cover AI mistakes?.

Practical controls

These steps reduce the risk:

See also AI-drafted assessments and care plans and predictive risk tools.

How Apex helps

Apex arranges PI for independent social workers and care consultancies working with councils on assessments and reviews. A broker can check how your policy treats work done with a client's AI systems; see PI for social workers and the AI in social work hub.

Sources

Frequently asked

Can an AI tool decide someone's eligibility for care?

A significant decision based solely on automated processing of health or disability data is restricted under UK GDPR Articles 22A to 22D, in force from 5 February 2026, and needs safeguards such as human intervention and the right to contest. In practice a social worker should consider the case and own the decision.

What should the record show if AI was used?

Standard 3.11 requires records that show how decisions were reached. Record the person's account, the evidence you weighed, your analysis and the outcome, and note where and why you departed from any tool suggestion.

Who is responsible if the tool is wrong?

The council is responsible for choosing and running the tool, and claims are usually handled through its arrangements. The social worker who signs off the decision remains professionally accountable for their judgement. SWE's research found this split is often unclear in practice.

Check your PI covers how you use AI

Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.

Start a PI quote Request a callback

Apex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.