Safeguarding culture or compliance: how to tell the difference

I have worked in international education since 2008, as an IB and MYP teacher, an IB DP economics teacher, an IB and CIS evaluator and workshop leader, and most recently as CIS accreditation coordinator at Miras International School in Astana. I am not a Designated Safeguarding Lead, and that matters for what follows, because the thing I noticed was administrative rather than clinical. The annual safeguarding training arrived as a standard module written for another country. It repeated almost unchanged from one year to the next, and a good number of staff passed it by remembering last year's answers.

The completion report at the end of that cycle was clean. Every name carried a tick and a date, and the file would have satisfied any visitor who asked for it. The practice behind the file was much thinner than the file suggested. That distance between a clean record and thin practice is the difference between compliance and a safeguarding culture in schools, and a Head can see it without commissioning anything, provided they look in the places where the two diverge.

A school running on compliance produces a recognisable picture. Completion sits at or near a hundred per cent. Certificates carry dates. The child protection policy was approved by the board in September and sits in the staff handbook. When an evaluator asks for the training record, the business manager sends it within the hour. Nothing in that picture is false, and nothing in it is wasted; a school that cannot produce records at all has a more urgent problem than the one described here.

The limit of the picture is what a record can describe. A completion row states that a member of staff opened a module and answered its questions correctly. It says nothing about what the same person does at four in the afternoon when a Year 8 student mentions something odd on the way to the bus, or what a cleaner does after noticing an adult alone with a child in a room that should have been empty. Culture shows up in those moments, and the record only catches them later, if at all.

The two states are not opposites. Every school with a strong safeguarding culture also keeps careful records, because the records are how the school checks itself and how an evaluator checks the school. The question a Head has to answer is narrower: whether the record is a by-product of practice, or a substitute for it.

Every sign below is observable inside a week, without a survey and without telling staff they are being assessed. Each one describes something a person does rather than something a document says.

• Staff name the Designated Safeguarding Lead and the deputy without looking at a poster, and can say where that person sits during the school day.
• The concerns log holds entries that came to nothing. A log with small, closed, no-further-action entries is stronger evidence than an empty one.
• Non-teaching staff appear in the log as reporters. Drivers, security officers, cleaners, catering staff and the school nurse raise concerns rather than only feature in them.
• The DSL receives questions in November as well as during induction week. Ordinary questions between formal reports show that staff treat the DSL as a person to consult rather than a form to submit.
• Someone other than the author has read the policy this year. Heads of department describe the reporting route in their own words and describe the same route.
• The board sees anonymised safeguarding data at least once a term, and asks about the pattern rather than the total.
• Records show what the school did after a concern arrived, including the decision not to escalate and the reason for it.

This costs about three hours of a Head's time spread across five days, and it produces evidence rather than impressions. Write the answers down as they come, in the words people use, because the phrasing carries as much information as the content.

• Monday: ask six members of staff, chosen across teaching, administration, facilities and transport, who the DSL is and what they would do if a student told them something worrying on the way home.
• Tuesday: open the concerns log for the last twelve months. Count entries, count distinct reporters, and count how many entries came from staff who do not teach.
• Wednesday: work through the most recent training module in the version staff actually saw. Mark every place it names a country, an agency, a helpline or a law that does not apply on this campus.
• Thursday: ask the DSL how many informal questions they received in the last month, and what the questions were about.
• Friday: read the last board pack. Note whether safeguarding appeared as a number, as a narrative with cases and decisions, or not at all.

If five of the six staff name the DSL and describe a route that matches the policy, the training is landing and the route is known. If the two who fail are the driver and the cleaner, the school has a coverage problem rather than a culture problem, and the fix is a training and language decision rather than a values campaign. If the teachers answer confidently and describe three different routes, the policy is either unclear or has been revised without anyone re-teaching it.

The log usually gives the sharpest reading. A school of four hundred students with two entries in a year is not a school where nothing happened; it is a school where small things did not reach the log. Concerns about adults tend to disappear first, which is why a written low-level concerns route for international schools is worth checking separately from the child protection route.

The module review on Wednesday is the one that tends to surprise people. Modules bought as generic courses routinely instruct staff to call agencies that do not operate in the country, cite thresholds set by another jurisdiction, and describe a reporting chain the school does not use. Staff notice the mismatch and quietly discount the whole module, including the parts that were accurate.

Three causes account for most of the gap I have seen. The first is training written for another school, which teaches concepts correctly and procedures wrongly, and which staff therefore treat as an exercise. The second is friction on the route: one named person, one channel, and no written option, so a concern about a colleague has nowhere quiet to go. The third is language. A school whose teachers work in English but whose drivers, cleaners, security officers and nurses think in Russian, Kazakh, Arabic or Turkish has trained half its adults and recorded all of them.

The response follows the causes in order. Rewriting the training against the school's own child protection policy is the fastest change and usually takes a term, because the material stops contradicting what staff are told in briefings. Giving the reporting route more than one entry point, including a written and a confidential one, changes the log within a year. Building the material in the languages staff use closes the coverage gap that the Monday conversations exposed. Schools that want the argument in full will find it in the comparison between generic safeguarding e-learning and school-specific training, and in the practical design of a reporting route that staff use when hierarchy discourages them.

Two documents are worth keeping open during that work. Keeping children safe in education, the Department for Education's statutory guidance for schools in England, sets out what a whole-school approach and a low-level concerns process are expected to contain, and many international schools use it as a reference even where it does not bind them legally. The International Task Force on Child Protection, whose materials sit on the ICMEC Education Portal, publishes expectations and self-audit tools written specifically for international schools. Neither one measures culture for you. Both give a Head language for describing what the school is aiming at.

Timeframes are worth stating plainly. The training can change in a term. The log usually changes in a year, once staff have seen a small report handled well and nothing bad happened to the person who made it. What people do when nobody is watching moves more slowly than either, and the annual refresher is where most schools lose the ground they gained, which is the subject of why repeated refresher training fails and what replaces it.

SafeguardIS builds a school's training from that school's own safeguarding and child protection policies, so the reporting route in the module is the route in the corridor. The Designated Safeguarding Lead reviews and approves every module before staff see it; publishing requires a named reviewer's declaration, and the approved content is hashed and logged, so a school can show later exactly what a member of staff was taught and who signed it off. Certificates are countersigned by the Head of School and the DSL and carry a code that anyone can check on a public verification page.

Two features address the gaps the self-assessment tends to expose. Staff can ask the AI policy assistant questions that are answered from the school's own policies, which gives the quiet member of staff a way to check a procedure without approaching a senior colleague; the DSL keeps every safeguarding decision, and no child's details belong in a chat window. A policy audit feature checks a school's documents against ITFCP-aligned expectations and produces an action plan. The platform ships training in English, Russian and Kazakh today. If your staff work in a language the platform does not ship yet, I build that language in for your school as part of the pilot. I also run live workshops in English and Russian and whole-school safeguarding audits, and Miras International School in Astana is the pilot school.

If you want to walk through the self-assessment above against your own records, book a 20-minute walkthrough and bring your training report and your concerns log.