New Ofsted AI Guidance: What Social Care Providers Need to Know

Artificial intelligence (AI) is increasingly present in education and social care settings, and now Ofsted is making its position clear. In June 2025, Ofsted published its guidance “How Ofsted looks at AI during inspection and regulation” (21 October 2025). GOV.UK
For providers working across children’s homes, supported accommodation, fostering, and other settings, this guidance is timely. Here’s what you need to know, and how you might respond.

Key themes from the guidance

1. AI defined, and the context
   Ofsted begins by defining AI as “an umbrella term for a range of technologies based on algorithms … designed to complete tasks that previously required human thinking.”
   It states that this is part of the broader government AI agenda, but emphasises this guidance is not a final position and will evolve.
2. Inspection/regulation approach
   • Ofsted supports innovation and the use of AI where it improves education and care of children and learners.
   • However: inspectors will not treat AI as a standalone item for inspection; they do not evaluate the tool itself, only the impact of its use in the setting.
   • Inspectors are not required to actively seek out AI use; nor will an inspection report automatically reference AI unless it has significant impact on children/learners.
3. Expectations for providers
   • Providers are not required to use AI.
   • But they should be aware that AI may be used by children, staff or across the setting (for example, homework support, case-note automation).
   • Inspectors may ask how leaders ensure AI use is in children’s/learners’ best interests. Example: AI used to identify causes of absence; or AI summaries of child-protection conferences.
4. Risks and how they are considered
   Although there’s no separate inspection domain for AI, the guidance highlights risk areas that intersect with existing inspection themes:
   • Data protection (AI often uses large data sets, possibly personal data)
   • Safeguarding: AI may introduce new risk vectors; inspectors will consider such risks as part of wider safeguarding evaluations.
   • Bias and discrimination: AI may perpetuate or amplify bias in its training data; providers should consider mitigation.
     The key question: What decisions did the provider make? What were their assurances? Inspectors won’t judge the algorithm; they will judge the provider’s decisions and their impact.

Implications for children’s homes, supported accommodation & staffing services

For your work in children’s social care settings (homes, supported accommodation, staffing), the guidance offers an opportunity and a caution. Here are practical implications:

• Governance & leadership: Leaders should be able to articulate whether AI is in use (by staff, by children, by the service), and if so: what decisions led to its introduction, how it is monitored, how risks (data, safeguarding, bias) are addressed.
• Policy & practice:
  • Review existing policies around use of AI (for example: children using chatbots, writing essays, staff using AI for notes or planning).
  • Ensure children-facing technology is aligned with best interest of young people, and staff understand how to respond if AI mis-used or prompts risk.
• Staff awareness and training: Make sure staff understand what AI might mean in your setting (e.g., generative text, predictive algorithms, automated summaries) and are prepared to ask critical questions: “Is this tool accurate? Who monitors it? What happens if it fails or is biased?”
• Safeguarding oversight: Since AI can shortcut human judgement or introduce unexamined bias, safeguarding leads should be aware of its use and ready to respond to issues like: misuse by young people (homework cheating? misuse of generative tools), data privacy concerns, algorithmic decisions affecting children.
• Data protection & compliance: If AI is processing personal data (young people’s records, staff notes, behavioural logs) you should treat it in line with GDPR/data-protection expectations. Be ready to explain to inspectors how you ensure data is managed, processed, stored, with appropriate safeguards.
• Inspection readiness: Although Ofsted won’t ask specifically about “AI” unless relevant, you should be inspection-ready:
  • Have a clear narrative of how you use technology/tools in your service and how you ensure they support children’s outcomes.
  • Be ready to show how you assess whether a tool has had a positive or negative impact.
  • Be able to show how you respond when tools don’t work as intended (mitigation, human oversight, review).

Practical checklist for services

Here’s a simple checklist you might use or adapt for your service:

• Map out any AI or AI-adjacent tools being used (by staff, by children, in the setting).
• For each tool, document: purpose, expected benefit, actual monitoring/evaluation, who is accountable.
• Review your policy framework to include reference to: use of generative AI; children’s access/use; staff use of AI in case-notes/planning; data privacy/processing.
• Ensure safeguarding/behaviour policies reflect potential AI misuse (cheating, automated content, bias).
• Provide staff awareness/training: what is AI, how it might appear in your setting, what risks exist, how to respond if something goes wrong.
• Ensure your leadership can articulate how you decide to adopt (or not) any AI tool, and how you assure its impact.
• Maintain documentation of decision-making and review cycles for any AI tool.
• Be ready for inspection: link technology use to children’s outcomes, have reflective evaluation of benefit/risk, be transparent.

Final thoughts

The Ofsted guidance reinforces a key message: technology (including AI) is only meaningful insofar as it supports children’s and learners’ outcomes. Ofsted is not looking for “Are you using AI?” but rather “If you use it, how and to what effect?”. In children’s social care it’s especially important that technology supports safe, ethical, life-changing work, not just efficiency for its own sake.

For providers in children’s social care, that means you can embrace innovation, but you need robust leadership, governance, policy, monitoring and a clear focus on the best interests of children and young people.

Full guidance at: How Ofsted looks at AI during inspection and regulation – GOV.UK