20 July 2026 · CareTime
Editor's note: Updated July 2026 with new AI adoption data and consumer survey findings on family attitudes to care home AI.
AI in care homes isn't a future concept — it's already here. Research published by homecare technology company Birdie in July 2026 found that 70% of UK homecare providers are now using AI in some form, with the figure projected to reach 85% within a year. This guide covers what's actually working in care homes right now, what's coming next, and how to evaluate whether any of it is worth your time.
The Birdie July 2026 research — drawn from the homecare (domiciliary care) sector rather than residential care specifically — gives the clearest sector-wide adoption picture published so far. Among providers who were re-inspected by CQC after adopting AI, 59% saw their CQC rating improve.
There is, however, a governance gap. Only 66% of AI-using providers have any formal governance policy. Around a third of organisations are running AI tools without a written framework for how the system operates, how outputs are reviewed, or how errors are escalated. Given the CQC's Well-Led emphasis on oversight and governance, this matters.
AI technology in the care sector falls into a few broad categories. Some are well-established, others are still emerging.
Call monitoring and intelligence. Care homes receive a high volume of incoming calls — from families, healthcare professionals, suppliers, and sales callers. AI call monitoring systems log incoming calls, screen numbers already flagged as nuisance callers, and surface what matters in a daily summary. CareTime's Silent Guard, for example, delivers an 8am Morning Brief that tells the manager what was flagged from the previous day — patterns of concern, vulnerable residents contacted, anything worth following up — so that the important calls get acted on rather than lost in the noise.
Fall detection and movement monitoring. AI-powered sensors can monitor resident movement patterns and detect falls in real time. Systems from providers like Vayyar and HoneyComb use radar or acoustic monitoring to alert staff without requiring residents to wear anything. These are particularly valuable for residents with dementia who may not be able to press a call button.
Acoustic monitoring. Sound recognition systems listen for specific sounds of concern — falls, distressed vocalisations, unusual coughing — and alert staff. Unlike camera-based systems, acoustic monitoring is less intrusive and raises fewer privacy concerns.
Care planning and compliance. AI tools are beginning to help with the administrative burden of care planning and CQC documentation. The key distinction is between AI that captures and organises real events from care interactions, and AI that generates generic text. Care England has explicitly warned that AI-generated CQC policy documents — too generic, possibly referencing outdated legislation, not reflective of actual staffing or local context — are a risk with inspectors. Tools that surface real call records and real events are a different category.
Staff scheduling and rostering. Predictive tools can analyse historical patterns to help managers build rosters that match demand, reducing both understaffing and overstaffing.
One area of genuine nuance for care homes considering AI phone systems: what families actually think. A survey reported by Care Home Professional found clear divides in consumer comfort depending on what the AI is being asked to do.
For administrative interactions, acceptance is fairly high: 40% of consumers would use AI to get general information from a care home (opening hours, visiting procedures); 33% would use it to book or rearrange a visit.
For personal conversations, the picture is very different: just 18% said they would be happy using AI to ask about the health or wellbeing of a loved one in care.
That gap matters when evaluating AI phone technology for your home. Families draw a clear line between administrative convenience and personal care conversations. AI systems positioned towards answering welfare enquiries are addressing exactly the use case families are least comfortable with.
AI receptionists for care homes. Natural language AI that can answer calls, handle family enquiries, book visits, and triage urgent calls. Several providers now offer care-home-specific answering services — including CareVoice AI, Softomate, and Moneypenny. CareTime's own AI Receptionist is on the waitlist for late 2026, built on top of the existing call monitoring platform. When evaluating answering services, the family comfort data above is worth bearing in mind: the administrative use cases (booking, basic information) are where acceptance is highest.
Predictive health monitoring. Wearable sensors and environmental monitors that can flag early signs of deterioration — changes in mobility, sleep patterns, or daily routines — before they become clinical events.
AI-assisted CQC preparation. Tools that continuously gather evidence of good practice from daily operations, making inspection preparation less of a scramble and more of an ongoing process.
Not every AI product is worth the investment. Some considerations:
Generic vs care-specific. Many AI tools are built for general business use and adapted for care as an afterthought. A generic call answering service won't understand the sensitivity of a family calling about their parent. Look for providers who have built specifically for the sector.
The governance gap. With roughly a third of AI-using care providers operating without a formal policy, this is a real risk — both operationally and from a CQC Well-Led perspective. Before deploying AI, have a written policy covering what the system does, who reviews its outputs, and how errors are escalated.
Data privacy. Any AI tool processing resident data needs to be GDPR compliant and ideally UK-hosted. Facial recognition and video monitoring raise particular concerns — elderly residents may not have capacity to consent to ongoing data collection. Always check where data is stored.
Integration with existing systems. The best AI tools sit alongside what you already have. If a provider needs you to change your phone system, care planning software, or hardware, factor in that disruption carefully.
Real evidence vs marketing claims. Ask for case studies and references from comparable care home settings. If a provider cannot show real-world results from the sector, that is a warning sign.
If you are a care home manager exploring AI for the first time, the lowest-risk starting point is a tool that doesn't require any changes to your existing setup. Call monitoring is a good example — it sits alongside your current phone system, provides immediate visibility, and demonstrates value within days.
CareTime's Silent Guard is available now as a 30-day pilot for £49 — no long-form contract, no hardware, no disruption. It is a practical way to see what AI call intelligence looks like in your home before committing to anything bigger. Get in touch.
AI in care homes is real and growing — 70% of UK homecare providers now use some form of it. The most useful tools focus on monitoring communications, improving visibility, and easing the administrative burden. There is a significant governance gap: only two-thirds of AI-using providers have a formal policy, and CQC's Well-Led framework will increasingly expect one. And families draw a clear line: comfortable with AI for admin, much less so for anything personal. The tools with the most immediate value are those that inform human decisions rather than replacing human conversations.
CareTime's Silent Guard is available now for a 30-day pilot. £49, 1-page pilot letter — exit by reply-email.
Join the 30-Day Pilot
CareTime