Home care technology in 2026: what agencies are actually using

Seven in ten UK homecare agencies are already using AI, and over half of them started before 2025. That is the finding from Birdie's survey of 122 UK homecare providers in spring 2026, and it reframes what home care technology means today. The question is no longer whether agencies will adopt it. It is whether they know what they have already adopted, and whether their governance covers it.

‍

This is what the research found about the technology homecare agencies are actually running, what it is doing to the quality of care they deliver, and what the gaps mean for the people running services.

‍

‍

What home care technology looks like in 2026

Most agencies now sit on a digital base. Care management platforms replaced paper care plans and paper notes across the sector over the past decade, and the digital social care record became the standard place where care information lives.

What has changed recently is the layer sitting on top of that base. AI has arrived, and it has arrived mostly from outside the care record.

Among agencies using or piloting AI, ChatGPT is the most common tool by a wide margin at 63%, followed by Microsoft Copilot at 48% and Google Gemini at 38%. General writing assistants such as Grammarly are used by 35%. Only 29% use AI built into their care management platform.

That distribution matters more than it first appears. The most common home care technology stack in 2026 is a care management system, plus a set of general-purpose consumer AI tools that have no connection to it.

Adoption also started earlier than the sector narrative suggests. Of the agencies using or piloting AI, more than half began before 2025:

‍

When agencies first started using AI-powered tools, as a share of AI users:

• Before 2022 - 7%
• 2022 - 9%
• 2023 - 19%
• 2024 - 16%
• 2025 - 33%
• 2026 - 16%

‍

Depth is a different story from breadth. Asked how intensively their teams use AI day to day, the most common answer was "used by some staff in some parts of the business" at 56%. Around a third, 34%, described it as deeply embedded with most staff using it regularly. So most agencies have crossed the line into using AI. Far fewer have made it the way the whole team works.

‍

‍

Agencies adopted it to improve care, not to cut costs

Homecare is one of the most financially pressured sectors in the UK, so the obvious assumption is that agencies reached for AI to save money. The data says otherwise.

Asked what prompted the decision to adopt, the most cited factor was wanting to improve care quality and outcomes, named by 62% of providers using AI. Staff time pressures and admin overload came second at 55%, and CQC inspection readiness third at 49%. Rising operational costs and margin pressure sat well down the list at 31%.

Where AI is being used follows the same logic. General office productivity is the most common area at 52%, but care planning and risk assessment runs almost level with it at 51%, ahead of compliance and quality management at 38%, care note writing at 31%, and medication management at 23%.

The second most widespread use of AI in homecare is a core care quality one. For around half the providers using it, AI is helping shape the care itself: the plans, the risk assessments and the records that decide what good support looks like for someone in their own home.

‍

‍

What that has done to care quality

Providers are convinced it is working. Asked whether AI has improved the quality of care they deliver, 76% said it had, with 42% saying it had improved significantly. A fifth, 21%, said quality had stayed the same, and 3% said it had declined somewhat.

The external benchmark points the same way. Among agencies re-inspected since adopting AI, 59% saw their CQC rating improve, and none reported a decline.

That number needs an honest caveat. CQC ratings move for many reasons and AI is one contributor among several, so this is correlation rather than proof. The CQC's own guidance on AI, published in May 2026, states plainly that "the absence or presence of AI does not predict a specific rating."

What makes the finding credible is the consistency across three separate measures: what managers perceive, how the regulator judges them, and where agencies choose to spend the time they get back.

‍

‍

Where the time saved actually goes

The operational gains are real, and they are larger than most agencies expect before they start.

Among providers able to quantify it, 67% estimate AI saves 15% or more of their office team's time on administrative and operational tasks. Almost four in ten, 38%, put it at 25% or higher. The most common answer, and the median, was a saving of 15–24%. A typical AI-using agency is now freeing up roughly a fifth of its back-office capacity.

The money picture is more mixed. A majority, 57%, reported a measurable reduction in operational costs, with 17% citing a fall of more than 15%. Around a quarter saw no measurable change.

The gap between hours saved and costs cut is the interesting part. Most agencies are not stripping out capacity to bank the savings. They are reinvesting it:

‍

Where agencies redirect the time AI saves, as a share of AI users:

• Improved care quality monitoring - 51%
• More direct time with care recipients or families - 48%
• Business development and growth - 41%
• Staff training and development - 31%
• Reduced overtime or out-of-hours working - 31%.

‍

The first financial dividend of AI in homecare is not a lower cost base. It is more capacity from the same team, usually without adding headcount.

‍

‍

The gap between the tools and the rules

Here is where the picture gets uncomfortable. Agencies moved before any rulebook existed, and the governance has not caught up.

Fewer than half of all agencies, 43%, have a formal written policy governing AI use. A third, 32%, have not formalised anything at all. Among agencies actively using AI, only two-thirds, 66%, have a formal policy, which means roughly a third are running AI in a regulated care setting with nothing written down.

Set that against what AI is being used for. Around half of AI users are applying it to care planning and risk assessment, and the tools doing the most work are general-purpose chatbots with no integration to the client record, no care-grade data handling and no audit trail for an inspector to follow.

Transparency with clients is thinner still. Among agencies actively using AI, 58% always inform clients and families when AI is used in their care or records, and a further 6% do so where it is particularly relevant. But 12% said they had not considered disclosure at all, and 5% said they did not think it necessary.

The regulatory position explains some of this, though it does not excuse it. The CQC published its first guidance on AI in health and social care in May 2026, and it is a principles-based document broad enough to cover everything from acute hospitals to residential care. As HTN reported in June 2026, sector-specific assessment frameworks are now in consultation. The Nuffield Trust called in March 2026 for national guidance, registries of approved suppliers and a formal strategic approach for social care.

Agencies know what they are missing. Asked what would most increase their confidence in AI, clear regulatory guidance from the CQC was the single most cited answer at 59%, ahead of AI tools built specifically for domiciliary care and better data-security guarantees, both at 43%. Asked how clear current guidance is on a nought to four scale, agencies returned a mean of almost exactly 2. More rated it unclear, 39% scoring nought or one, than clear, 35% scoring three or four, and the split is polarised at both ends: one in five called it "not at all clear" while 24% called it "very clear". When a sector cannot agree on whether useful guidance exists, no shared standard has taken hold.

‍

‍

Technology adoption is splitting the sector by size

The benefits are not reaching everyone. The clearest dividing line is scale.

Among the smallest agencies, those supporting fewer than 20 clients, just 15% are actively using AI. That rises to 43% of agencies with 21–60 clients, 48% of those with 61–140, and around 80% of agencies supporting 141 or more. The same pattern holds by footprint: 71% of multi-branch agencies are active users, against 41% of single-branch operators.

The reason is not price. Among agencies not yet using AI, the most common barriers were regulatory uncertainty and teams not feeling ready or skilled enough, both at 28%, followed by not understanding what AI would do for them and not knowing which tools to choose, both at 25%. "We can't justify the cost right now" sat near the bottom at 17%. This is a smaller group of respondents, so read those figures as directional.

Adopters describe the same problem from the other side. Their biggest challenge is integrating AI with existing systems, at 48%, followed by getting staff to trust the tools at 43% and inaccurate outputs needing correction at 34%.

Both groups are asking one question: where is the AI that was built for care, fits how we already work, and can be trusted with the people who depend on us?

‍

‍

What agencies want home care technology to do next

The direction of travel is towards care itself rather than the admin around it.

Asked where they most want to use AI over the next two years, agencies named predictive risk identification, meaning flagging clients whose health may be deteriorating early, and automated compliance and inspection readiness, each at 58%. AI-generated care notes and care plans followed at 56%, workforce planning and staff retention at 48%, and AI-assisted scheduling and rostering at 47%. Only 7% expressed interest in none of it.

Investment intent splits along the same adoption line. Across all agencies, 43% plan to increase investment in AI or digital tools over the next 12 months. Among those already using AI, that rises to 60%. The agencies with direct experience are committing more, not less.

‍

‍

‍

‍

‍

How to choose home care technology now

If you are reviewing your technology this year, the research points to five practical checks worth running before you commit to anything.

Does the AI sit inside the care record, or beside it? Integration is the single biggest reported challenge. AI bolted on from outside cannot see the client record, which limits what it can do and leaves no trail linking its output to the care it shaped.

Can you see where every output came from? If a suggestion cannot be traced back to the source, nobody can verify it, and an inspector cannot audit it.

Does a human approve before anything is saved? Suggestions that apply themselves are a governance problem. Suggestions a practitioner reviews and confirms are not.

Was it built for home-based care? Frameworks and tools designed for acute or residential settings and then applied to care in someone's home do not fit the work. Ask specifically about domiciliary care.

Do you have a written AI policy covering it? Data handling, accuracy checks, human oversight and disclosure to clients. If it is not written down, it will not hold up at inspection.

That last one is the quickest win available to most agencies, and the one a third of AI users are currently missing.

‍

If you are ready to compare platforms, start with this complete guide to domiciliary care software.

‍

Common questions about home care technology

What is home care technology?Home care technology is the software and hardware domiciliary care agencies use to plan, deliver, record and evidence care. In 2026 that usually means a care management platform holding digital care plans and notes, electronic medication records, scheduling and rostering, finance and invoicing, and increasingly a layer of AI tools that support care planning, note writing and compliance.

How many home care agencies in the UK use AI?Seven in ten. Birdie's 2026 survey of 122 UK homecare providers found 70% were using or piloting AI, a figure set to rise to 85% within a year. More than half of those using it started before 2025.

Does AI improve CQC ratings?Among agencies in the survey re-inspected since adopting AI, 59% saw their rating improve and none saw it decline. That is an association rather than proof of cause, and the CQC states that the presence or absence of AI does not predict a rating.

Is AI allowed in care under CQC rules?Yes. The CQC published its first guidance on AI in health and social care in May 2026, setting out principles rather than prohibitions, and sector-specific assessment frameworks are in consultation. Providers are expected to govern AI use, which means documented rules on data handling, accuracy, human oversight and disclosure.

What is a digital social care record?A digital social care record is the electronic version of a person's care plan and care notes, replacing paper files. It is the base layer most other home care technology connects to, and it is what any AI tool needs access to in order to be useful for care rather than just admin.

‍

‍

The full research

The findings above come from Moving faster than the rules: AI, care quality and the homecare sector in 2026, the first study to measure AI adoption, outcomes and governance in UK domiciliary care. It is based on 122 UK homecare providers surveyed in spring 2026, and it is free to read.

Read the report

‍