At the start of this year, I visited Downing Street alongside a social worker, Joanna. The PM was interested in how AI is driving social impact, and how it can be scaled.
No amount of clever tech or demos could have had the same impact on Starmer as Joanna’s first-hand account. She, and social workers like her in councils, spend more than three days of every working week on admin and paperwork.
Whilst a portion of this admin is high value – for example using professional expertise to draw insights and make recommendations for an individual’s care – the sad truth is that most of this time is spent simply transcribing notes from meetings and formatting them into reports.
Clearly, this is incredibly time consuming and financially costly.
But, more than that, I hear time and time again from social workers that it also means they struggle to deliver the best possible care. That the relentless need to manually keep records keeps their heads in a notebook, rather than focused on the person in front of them. That they spend long hours catching up on admin.
AI gives us a huge opportunity to change this.
The AI opportunity for social care
At Beam, we’ve been working with local authorities for more than seven years, directly delivering services that help homeless people into jobs and housing. Through that, we’ve achieved more than 5,000 outcomes.
Our team of over 100 caseworkers do amazing work. But, just like Joanna, they were held back by paperwork. We built Magic Notes to support them. Magic Notes works by recording and transcribing a social workers’ meeting with a service user, and then uses AI to generate casenotes, assessment forms, letters and any other output the social worker needs.
Its extraordinary impact led to us offering this technology to local authorities up and down the UK – and today more than a third of social care teams are using AI.
Through multiple independent evaluations, conducted by both Local Authorities and an academic team headed by Professor Rob Procter of the University of Warwick, we have seen third party evidence that Magic Notes can save each frontline worker more than 8 hours per week in admin time: that’s one full day a week.
Not only is that a huge productivity gain, but the AI is also enabling interactions to feel more human. Service users report that they prefer their interactions with social workers, because the social worker can look up at them and truly listen, without their head buried in a notebook scribbling notes.
This all means higher quality, more effective care. And its success is the reason Magic Notes is now rolling out to other settings, like prisons and hospitals.
AI adoption needs caution – especially in the care sector
But of course, the use of AI requires caution. In the care sector, especially, we need to be laser focused on privacy, security and safety.
Security is clearly of paramount importance when handling sensitive data. It’s good that security accreditations exist, and even better that the government is creating fertile ground for British tech companies, so data hopefully won’t ever have to cross the Atlantic for British citizens to receive world class services.
Beyond technical security, we must also consider how we use AI responsibly in this setting.
At Beam, we invest heavily in ensuring our AI tools produce clear, accurate outputs. We design our products so that a qualified, human worker has to check every output. We seek constant feedback from frontline workers. And our organisation has people with direct frontline experience at every level – from our co-founder, to our engineers writing code and prompts, to our sales team working with customers.
Consent and privacy is also baked into our product, and will be a vital component of the future of AI in care. Patients and service users must always be able to consent to their data being used and collected, and in particular understanding how it is used.
If we get this right, with responsibility at the core, we’ll fundamentally transform the nature of care. We’ll make jobs in healthcare, social work and justice more rewarding and meaningful, while delivering a better quality of care for service users. And that impact can be global.
