New research finds strong support for AI to ease social workers’ workloads

Social Work England has released two major research reports examining the emerging use of AI in social work education and practice in England. Together, the studies explore how AI is currently being used, the opportunities and risks it presents, how prepared the workforce feels, and what the implications may be for professional standards.

The research, carried out in spring 2025, found that generative AI is now the most commonly used form of AI in social work. Social workers, students and academics reported using tools such as virtual assistants, transcription software, case recording support and chatbots. However, take-up remains uneven across the sector. Some employers are actively embracing AI and offering training and guidance, while others are not using it at all because of cost, limited infrastructure or uncertainty about how best to deploy it.

Despite this variation, attitudes towards the potential benefits of AI were largely positive. In a survey of 155 social workers conducted as part of the research investigation by Research in Practice, 83% of respondents said AI had the potential to reduce administrative burden. A further 70% felt there were “equal or more benefits to using AI in social work than there are risks”.

Sarah Blackmore, Executive Director of Professional Practice and External Engagement at Social Work England, said: “As artificial intelligence continues to develop quickly, it is important that we understand how AI is being used in social work. These findings will help us understand the potential impact of using AI on public safety, public confidence and public trust in the social work profession”.

The findings suggest that AI tools are already being used to support note-taking, generate documentation and transcribe meetings, helping some practitioners reclaim time for direct work with people using services. Respondents also highlighted potential benefits around improved accessibility, reduced costs and enhanced workforce wellbeing, with some linking time saved on administration to lower stress and cognitive load.

However, optimism about AI was more cautious when it came to its role in decision-making. Fewer than half of respondents believed AI had the capacity to aid decision-making (48%) or to identify risk and need in social work (46%), underlining the continued importance of professional judgement.

Alongside the potential gains, both studies raised serious ethical and practical concerns. Key risks identified included privacy and data protection, consent, bias, accuracy, reliability and accountability. Researchers found examples where personal information had been shared with publicly available generative AI tools, despite uncertainty about how such systems store or process data.

The studies also highlighted dilemmas around informed consent, particularly when working with vulnerable people or those who may lack capacity. While consent to record conversations for transcription was seen as relatively straightforward, using AI more broadly to process or analyse information raised more complex questions.

Concerns about bias and discrimination featured prominently. The literature review conducted by The Open University warned that AI systems can amplify existing inequalities in data, potentially leading to unfair outcomes for marginalised groups. Both reports also questioned who should ultimately be accountable for AI-generated outputs, particularly where such tools influence case recording or recommendations.

Sarah Blackmore said: “We welcome the recommendations set out in the research and will consider how best to respond where they relate to our role in public protection and regulation. We hope those working across the social work sector find these reports useful as they develop their own approaches to using AI in social work settings. We also hope social workers continue to consider how they use AI so that this can be done in a safe and ethical way that enhances, not replaces, their practice and expert knowledge and skill, and does not put the public at risk.”

One of the most striking findings relates to education and training. The research found that 86% of social workers who graduated in the past five years did not receive any specific preparation on using AI in social work practice during their education. Engagement with AI among social work educators was also found to vary widely, with some providers actively exploring its potential and others focusing mainly on issues of academic integrity.

The reports argue that digital literacy and ethical use of AI must now become core components of social work education. Potential benefits in education include personalised learning, improved administrative efficiency and better preparation for future workplaces. However, educators reported a lack of clear guidance and support around how AI should be used responsibly in teaching and assessment.

The Open University’s literature review, which analysed 119 academic articles and 44 pieces of grey literature, found a shortage of robust empirical evidence specifically focused on AI in social work. Most available research relates to health and social care more broadly, highlighting the need for further study as AI technologies continue to evolve.

Both sets of researchers emphasised that ethical practice, governance and regulation must underpin any expansion of AI use in social work. They recommended an ongoing focus on critical thinking, professional judgement and decision-making, alongside continuing professional development to keep pace with rapid technological change.

The reports also call for collective leadership across government departments, regulators, professional bodies and employers to shape the lawful, ethical and responsible use of AI. They suggest that Social Work England should consider whether updates to its professional standards and guidance are needed to reflect the growing role of AI, and that people with lived and learned experience of social work should be involved in shaping future policy.

Social Work England says it will carefully consider the findings and recommendations and work with partners across the sector to support positive developments.