LocalignFor education and training

One safe AI environment for education and training.

One managed workspace where the school itself decides which models, which guardrails and which age limits apply. Pupils safely use multiple models, minors' data stays under GDPR, and model comparison itself becomes a piece of AI literacy.

Managed environmentAge limitsGDPR for minorsMulti-model
Read more
Safe learning, world-class AI
For whomPrimary educationSecondary educationVocational educationTrainersTraining providers
Try example
app.localign.ai/trial

About us

Local and aligned

Localign takes its name from Local and Alignment. Our mission is to bring world-class AI to organisations, on their own terms. For education that begins with one hard guarantee: minors' data is processed exclusively within the EU.

Most tools are consumer chatbots without oversight, which the school cannot control. Localign processes the work by default with an internal, EU-hosted model and consults external models only anonymised, through a security shield. That way pupils safely use multiple top models while the school decides the models, guardrails and age limits and data stays under GDPR.

The context

AI is already part of education. The question is: who keeps control.

Teachers and pupils have been using AI for a while, often on separate consumer chatbots without oversight.

Minors' data leaks to American AI services, out of sight of the data protection officer.
The school has no grip on which models pupils use and what they do with them.
No age limits, no guardrails and no overview for teacher or board.

The opportunity

One managed environment for the whole school, instead of separate chatbots.
Minors' data under GDPR, within the EU, demonstrably.
Models, guardrails and age limits that the school decides itself.
Safe, equal access for every pupil, regardless of whether there is a paid account at home.

What teachers and pupils do with it

Safe learning and teaching, with control

The first gains are in everyday teaching work: creating lesson material, setting tests and giving feedback, and letting pupils practise safely with AI.

Generate and differentiate lesson material

Create lesson material at the right level, with variants for different pupils, based on the school's own method.

Test questions and feedback

Generate test questions and clear feedback per mistake, aligned to the learning goals.

Shielded pupil access

Let pupils safely use multiple models, within the school's guardrails and age limits.

Model comparison as pedagogy

Let pupils compare and critically assess models, as an exercise in AI literacy.

Summarise and explain

Long texts reduced to the essentials and explained in plain language, at the right level.

Translate and simplify

Make lesson material accessible for non-native pupils and at different language levels.

In principle this is possible with the large American providers too. The difference: with Localign it is a managed environment, minors' data stays under GDPR within the EU, and the school keeps control: the teacher reviews and decides, AI delivers drafts.

It is this simple

You ask. Localign arranges it.

One instruction, a team of AI assistants. Localign understands your question, distributes the work across the right assistants and delivers a single result, with pupil data protected and the school's guardrails at every step.

01

Understands your instruction

splits it into sub-questions
Your instruction

Make three versions of this practice test on fractions for class 3A, at the right level, with feedback per mistake.

02

Distributes the work across assistants

each model, each the right limits

Internal model

Own method

Which learning goals and practice material belong to this chapter on fractions in our method?

Anonymised external model

Differentiation

Which explanation suits Sara, dyslexia?

Pupil data removed

Which explanation suits a pupil with dyslexia?

ClaudeMethod

Guardrail active

Age limit

Which models and filters apply to this class?

Age limit active

Access and models tuned to:

class 3A · ages 14-15
AdjustApprove
03

Merges into one result

in the school environment
Differentiated test

Bundles the learning goals, the differentiation and the feedback into three versions of the test. The teacher reviews and decides; Localign delivers the draft. Pupil data is protected and everything stays within the school's guardrails.

Audit log
Detected · pupil dataSara, class 3A
Removed for external modelreplaced by 'a pupil'
Blocked beyond age limitunsuitable model excluded for this class
Recorded for the schoolvisible to teacher and board

Control

The school decides which models pupils use

The work is processed by default by an internal, EU-hosted model. Switch each external model on or off per class or subject; a security shield masks pupil data before every external model.

By default an internal, EU-hosted model processes the work.
Set per class or subject which models are on, within age limits.
Pupil data is masked before every external model, through a security shield.

Available models

These are the models pupils and teachers can use. The school decides which are on and which guardrails and age limits apply; sensitive data is masked before sending.

O

OpenAI

ChatGPT

A

Anthropic

Claude

X

xAI

Grok

G

Google

Gemini

M

Mistral

Mistral

P

Perplexity

Sonar

On your terms

Three promises your school can build on

Management

The school at the wheel

You decide which models, guardrails and age limits apply, per class and per subject. With insight into usage and a full overview for teacher and board.

Safety

Minors' data under GDPR

Pupil data is processed exclusively by European companies within the EU, not used to train models, and masked before every external model.

Literacy

Model comparison as pedagogy

Pupils safely use multiple models and learn to compare them critically. That way the platform itself becomes a piece of AI literacy.

Built for safe education
GDPREducation privacy covenantSIVON AI assessment frameworkIBP standards frameworkEU AI ActDPIA support

Who stands behind it

Built by people who know education

Localign was founded by people who know education and safety from the inside, from the classroom to cybersecurity.

Casper Rutjes

Casper Rutjes

CEO and co-founder

Dr. Casper Rutjes is a qualified secondary-school physics teacher (via the Eerst de Klas programme) and enjoyed his time in the classroom. He also led AI transformations as CTO at Amsterdam Data Collective, focused on responsible AI.

Mathyn Scheerder

Mathyn Scheerder

CTO and co-founder

Mathyn Scheerder worked as Tech Lead at the National Cyber Security Centre (NCSC) on protecting national digital infrastructure, and built secure data platforms as a data engineer.

Frequently asked questions

What happens to minors' data?

Pupil data is processed exclusively by European companies within the EU and not used to train models. Only masked or rewritten text goes to external models, and every step is in the audit log.

Does Localign align with the SIVON assessment framework and the IBP standards framework?

Yes. Localign is set up to align with the SIVON AI assessment framework and the IBP standards framework, and provides the documentation and DPIA support you need. Transparency, human oversight and logging align with the requirements of the EU AI Act.

Can the school set age limits and guardrails?

Yes. Per class or subject the school decides which models are on, which topics are shielded and which age limits apply. Localign helps enforce the generative-AI school agreements from Kennisnet and the VO-raad, with an overview for teacher and board.

Does Localign work with our login and learning environment?

Yes. Logging in works with your existing accounts and with single sign-on via Entree Federatie. Localign connects to your learning environment or student information system, such as Magister or Somtoday. No installation and no managing separate models or API keys.

Does procurement go through SIVON or the school board?

Either works. Localign aligns with the agreements of SIVON and the Education Privacy Covenant, and can also be purchased directly via the school board.

Doesn't AI undermine learning and assessment?

Not when it is under the teacher's direction. Localign is aimed at learning and AI literacy: pupils compare and critically assess models, the teacher decides where AI may and may not be used, and the audit log makes usage transparent. That way the teacher keeps watch over assessment and academic integrity.

Does the AI take over teaching or grading?

No. Localign delivers drafts, practice material and feedback; the teacher reviews and decides. The teacher stays responsible and every step is traceable via the audit log.

Ready for one safe AI environment at school?

Book a 30-minute demo or request the privacy documentation. We show how Localign fits within the school's requirements, using your own lesson material.