The Department for Education has published Keeping Children Safe in Education 2026, the statutory safeguarding guidance for schools and colleges in England.

An important note before going any further: this version is currently published for information only. It does not come into force until 1 September 2026, when it will replace the September 2025 guidance. The document itself states that a final version will be published on that date. The wording could therefore still change, and schools and colleges must continue to follow the current 2025 guidance until then.

Nevertheless, the version published in advance gives educators, leaders, governors and indeed all school and college staff much to consider before the new academic year.

There are substantial developments across the document. These include changes relating to harmful sexual behaviour, artificial intelligence, online safety, mobile phones, information sharing, alternative provision, reasonable force, safer recruitment, mental health and the specific barriers experienced by children with SEND.

One theme, however, particularly stood out to me: a clearer emphasis on support beginning before statutory intervention is required.

Support before statutory intervention

Part One now includes the heading Support before statutory intervention. Paragraph 17 states that any child may benefit from support before statutory intervention, including support from universal services, community – based Early Help or the targeted early – help level of Family Help.

This appears to describe a continuum rather than a single threshold:

universal support → community – based Early Help → targeted Family Help → statutory intervention

The distinction matters.

Safeguarding can sometimes be understood primarily as the action taken once significant harm is suspected or a referral threshold has been reached. Those actions remain essential. However, the 2026 guidance also directs attention towards the help children and families may need while difficulties are emerging.

All staff should understand the relevant local processes, criteria and routes for support. Annex C confirms that the guidance has been revised to distinguish more clearly between community – based Early Help and targeted early help delivered through Family Help.

Earlier involvement is therefore not simply about making referrals sooner. It can include noticing emerging barriers, drawing together information, listening to the child and family, strengthening universal provision, making appropriate adjustments and involving other practitioners before difficulties become entrenched.

Why this matters for children with SEND

The connection with SEND is explicit.

Under the new heading, staff are reminded to remain particularly alert to children who are disabled, have particular health conditions, have special educational needs – with or without an Education, Health and Care Plan – or have a mental health need.

The guidance also recognises further barriers that children with SEND may experience when abuse or neglect occurs. Their experiences may be wrongly attributed to their disability or SEND. They may face communication barriers when trying to express a concern. Behaviour, mood or injury may be explained too quickly as part of an identified need rather than explored with appropriate professional curiosity.

This does not mean that every SEND concern is a safeguarding concern. Nor should ordinary difference, distress or dysregulation be pathologised as evidence of abuse. It does mean that SEND, wellbeing and safeguarding cannot safely be treated as entirely separate conversations.

A child may be unable to communicate what is happening in the expected way. They may not recognise an experience as harmful or know how to disclose it. They may have learnt that adults interpret their distress as challenging behaviour. They may mask difficulties in school and release their distress elsewhere. Trusted relationships, accessible communication and careful attention to changes in presentation are therefore central to effective safeguarding.

When the outcome is mistaken for the problem

The expanded list in paragraph 17 is especially significant. It includes children who:

  • are frequently missing from education, home or care;
  • have repeatedly been removed from the classroom;
  • have experienced multiple suspensions;
  • are on a part-time timetable;
  • are at risk of permanent exclusion; or
  • are being educated in alternative provision or a pupil referral unit.

These experiences may be treated separately as attendance, behaviour or placement issues. KCSIE 2026 asks staff to consider whether they may also indicate a need for additional support before statutory intervention.

That does not remove personal responsibility or mean that every instance has the same explanation. It does invite professional curiosity.

  • What has happened before a pupil stopped attending?
  • What barriers have made participation increasingly difficult?
  • What is repeated removal from lessons communicating about the pupil, the environment or the interaction between them?
  • Has a part-time timetable become a holding arrangement without a clear route back to sustainable education?
  • Have support and assessment developed alongside one another, or has the child been expected to deteriorate sufficiently to meet a service threshold?

In my recent article, Attendance Is an Outcome, Not a Behaviour: A Reflection on Targets, Belonging and SEND, I argued that attendance data tells us what is happening but not necessarily why. Absence may be the eventual outcome of anxiety, unmet need, weakened belonging, inaccessible environments or repeated experiences of failure.

KCSIE 2026 adds an important safeguarding dimension to that discussion. Repeated absence, exclusion and placement instability are not merely outcomes to be corrected. They may be warning signs requiring curiosity, information sharing and coordinated support.

Participation before diagnosis

This also connects closely with Participation Before Diagnosis: Reflections on Inclusive Teaching, Engagement and the Interim Review.

In that article, I argued that schools should respond when barriers to access, participation and wellbeing become apparent rather than waiting for diagnosis, prolonged failure or crisis. Diagnosis can be valuable and important, but it should not be the point at which adults first begin to understand or support a child.

KCSIE uses different language and serves a different purpose. It is safeguarding guidance, not a SEND framework. Nevertheless, the underlying direction feels compatible:

Need should initiate support; a threshold should not initiate concern.

Schools do not need to diagnose autism, ADHD or another condition before increasing predictability, adapting communication, investigating sensory or relational barriers, strengthening trusted relationships, gathering pupil and family views or reviewing whether participation is sustainable.

Similarly, earlier universal or targeted support should not be used as a reason to withhold specialist assessment or intervention where these are required. Early support must mean that appropriate help begins sooner – not that children receive less.

Earlier involvement must also mean doing with

There is another necessary qualification. Intervention can happen early and still be experienced as something imposed upon a child.

A school might respond quickly by increasing monitoring, introducing a behaviour chart or placing responsibility on the pupil to regulate themselves more effectively. Speed alone does not make an intervention inclusive, relational or helpful.

In From Doing To to Doing With: Rethinking Regulation Support for Neurodivergent Pupils, I considered the difference between support designed for or imposed on a child and support developed with them.

Doing with means beginning with curiosity, taking communication seriously, protecting dignity and agency, considering what adults and environments can change and involving the young person in deciding what helpful support looks like.

This feels particularly relevant to KCSIE’s emphasis on child-centred practice. The guidance reminds staff that children may not feel ready or know how to tell someone what is happening. They may not recognise an experience as harmful. Staff should retain professional curiosity while also building trusted relationships that make communication possible.

Listening is not an optional addition after adults have made all the important decisions. It is part of understanding risk, need and effective support.

Two practical tools for beginning the conversation

Guidance establishes expectations, but staff may still ask what earlier, relational support looks like in practice. Two resources I have developed may offer useful starting points.

My Getting Back to Calm Guide: Printable Scripts and Tools accompanies the article Getting Back to Calm: Co – Regulation and Restorative Conversations in Upper Key Stage 2. It separates support into what adults can do before a flashpoint, during distress and after the event. Its emphasis is on calm adult presence, co-regulation, privacy, pupil voice and restorative – not punitive – follow-up.

Although it was written with upper Key Stage 2 in mind, many of its principles can be adapted thoughtfully for other ages and settings. The guide is not a safeguarding procedure and should never delay recording or reporting a concern. It is a practical aid for responding in ways that protect safety, dignity and connection while the school follows its established safeguarding processes.

The PERMA(H) Check – In Proforma was designed as a calm, low-pressure conversation for a young person who is not currently attending school. It explores positive emotion, engagement, relationships, meaning, accomplishment and health, while asking what helps, what feels difficult and what adults or school could change.

The proforma allows questions to be paused or skipped, makes next steps optional and includes space for pupil voice and helpful adult action. This makes it potentially useful when absence or disengagement is emerging and adults need to understand the young person’s experience rather than moving immediately to targets or compliance.

Again, it is a conversation scaffold – not a diagnostic assessment, safeguarding risk assessment or substitute for referral. If a safeguarding concern arises, staff must follow their setting’s procedures. Used within those boundaries, it may help staff build the trusted relationship, accessible communication and fuller understanding that effective early support requires.

A responsibility shared by all staff

KCSIE 2026 also removes the shortened version of Part One that previously appeared as Annex A. The advance guidance states that all staff – including staff who do not work directly with children – should read Part One in full.

That reinforces the idea that safeguarding is not confined to the designated safeguarding lead or senior leadership team.

Children reveal difficulty in many different ways and to many different people. A teaching assistant may notice a change in communication. A lunchtime supervisor may observe isolation or a shift in peer relationships. An administrator may recognise a pattern in messages from home. A site or transport colleague may see something that is not visible in lessons. A class teacher may notice that a previously engaged pupil is increasingly absent, distressed or unable to participate.

No individual member of staff holds the whole picture. A safeguarding culture depends upon people recognising what may matter, recording it appropriately and sharing it through the agreed process.

For SEND professionals, SENDCos, pastoral teams and advisory services, this also strengthens the case for close working with designated safeguarding leads. Behaviour, attendance, communication, mental health, family circumstances, participation and safeguarding may overlap. Effective support requires those pieces to be considered together without assuming that any single explanation accounts for everything.

What might schools consider now?

Because this guidance does not come into force until 1 September 2026 and may still change, schools should avoid treating the advance version as the final text. It does, however, provide a useful opportunity for reflection and preparation.

Questions for schools and colleges might include:

  1. Do all staff understand the difference between universal support, community-based Early Help, targeted Family Help and statutory intervention?
  2. Are staff confident about local referral and assessment routes?
  3. Do SEND, safeguarding, attendance, behaviour and pastoral systems communicate effectively with one another?
  4. Are repeated absence, removal, suspension or part-time timetables prompting curiosity about unmet need as well as action on the presenting outcome?
  5. Can children communicate concerns in forms that are accessible to them, including through writing, visuals, AAC, movement, behaviour or trusted adults?
  6. Are pupil and family views considered early, or mainly once a situation has escalated?
  7. Does support examine environmental and systemic barriers as well as the child’s skills and behaviour?
  8. Is specialist advice sought while there remains scope to prevent crisis, exclusion or placement breakdown?
  9. Do records distinguish clearly between observation, professional interpretation and verified information?
  10. Are staff prepared for everyone to read Part One in full from September?

A developing direction

KCSIE 2026 does not provide a complete model of inclusive education, nor should safeguarding language be used to turn every educational difficulty into a child-protection issue.

What it does appear to offer – as currently drafted – is a clearer recognition that effective safeguarding includes noticing vulnerability, responding to emerging need and coordinating support before statutory intervention becomes necessary.

Taken alongside wider developments in inclusive teaching, attendance and neurodiversity, this contributes to an increasingly consistent direction:

  • do not wait for diagnosis before addressing visible barriers;
  • do not mistake absence, exclusion or dysregulation for the whole problem;
  • do not impose support without understanding the child’s experience;
  • do not wait for a situation to become a crisis before bringing people together.

Perhaps the central question for schools is no longer simply:

Has this child met the threshold for intervention?

It is also:

What can we understand and do now to prevent their needs, vulnerability or exclusion from escalating?

That is not about lowering thresholds indiscriminately. It is about ensuring that children do not have to deteriorate sufficiently to reach them before meaningful support begins.

This article reflects the advance version of Keeping Children Safe in Education 2026 available on 10 August 2026. The guidance is due to come into force on 1 September 2026, when the Department for Education has stated that a final version will be published. Readers should check the final statutory guidance before using this article for policy or training purposes.