AI should not sit beside the school platform with no understanding of the work around it.
EdiWay AI is designed to support authorised users inside the learner, teaching, leadership, family, SEND, workforce and operational workflows they already use.
Ask questions. Prepare drafts. Summarise selected information. Organise evidence. Reduce repetitive writing. Support planning.
Then keep the important decisions with people.
Different school roles need different kinds of assistance. EdiWay AI can support appropriate workflows across areas such as:
Lesson support, resource preparation and feedback drafting.
Data questions, reports, summaries and evidence preparation.
Draft messages for parents, carers, learners and staff.
Help organise authorised information and prepare working drafts.
Summarise selected evidence and prepare review material.
Draft agendas, questions, summaries and follow-up communication.
Help organise evidence, policies and working reports.
Support appropriate staff-development and administrative drafting. The AI should work with the user’s permitted context. It should not become a separate unrestricted copy of the school record.
One of the most important principles behind EdiWay AI is that AI access should follow the platform's permission model. An AI request may need to consider:
Who is asking?
Which organisation are they working within?
Which learner, if any, is in scope?
What responsibility does the user hold?
Why does that person have access to this learner?
Is the relevant information teaching, SEND, safeguarding, medical, HR or another restricted area?
Why is the information being used?
Is the user authorised to access it? EdiWay AI is designed to inherit access boundaries rather than bypass them. A teacher using AI should not automatically gain access to information that the teacher could not otherwise view.
EdiWay's AI foundations include natural-language assistance and school-data search. Where enabled and appropriately permissioned, this can help authorised users ask questions about information available within their role. For example:
Help locate relevant permitted records or resources.
Prepare a working summary from selected information.
Make complex school-held information easier to review.
Bring relevant points together for a meeting or task.
Help structure permitted factual information for human interpretation.
Create a working starting point for further action. AI answers should support investigation. They should not replace checking the underlying records where accuracy matters.
Teachers spend significant time preparing, adapting, reviewing and communicating. EdiWay AI can support appropriate teacher workflows such as:
Help prepare working lesson content.
Suggest or organise teaching-resource material.
Help structure learner tasks.
Prepare feedback for teacher review.
Organise selected information relevant to teaching.
Draft appropriate parent and carer updates.
Help turn permitted evidence into a working narrative. AI can provide a starting point. The teacher remains responsible for educational suitability, factual accuracy, feedback and final judgement.
EdiWay's resource and lesson foundations can connect AI assistance with authorised teaching context. That may include:
Prepare working teaching material.
Use appropriate class or lesson information where enabled.
Suggest possible ways to adapt material for teacher consideration.
Generate working activity ideas.
Assist with draft responses to learner work.
Suggest possibilities for further learning. Generated material should always be reviewed before use. EdiWay AI does not certify that a generated resource is accurate, accessible, curriculum-aligned or suitable for every learner.
EdiWay AI can assist authorised staff with communication drafting. That may include:
Prepare clear working drafts.
Help make information age-appropriate.
Structure internal communication.
Draft preparation information.
Turn agreed human actions into a working communication.
Prepare supported translated drafts where enabled. AI does not determine who is entitled to receive sensitive information. Before sending, the authorised user remains responsible for checking the facts, recipient, tone, permissions and context.
Reporting can involve repeatedly bringing together information already held elsewhere. EdiWay AI can help authorised users prepare working drafts such as:
Organise selected permitted information.
Bring relevant evidence together.
Prepare narrative around approved information.
Structure the important points from authorised material.
Help organise selected records.
Turn approved information into clearer working language. A generated summary is not automatically the authoritative record. Users should return to the source information where decisions depend on accuracy.
SEND work can involve large amounts of information from different people and stages of the learner journey. Where enabled, EdiWay AI may assist authorised users with:
Organise selected learner information.
Prepare working summaries for human review.
Suggest possible strategies for professional consideration.
Suggest possible support approaches.
Help structure school-held working material.
Assist with drafting clear support information.
Help identify questions or information requiring review. AI assistance remains separate from professional and statutory judgement.
AI may help teachers work with assessment information. That can include:
Prepare possible feedback.
Organise selected learner evidence.
Draft working explanations.
Suggest areas a teacher may wish to consider.
Help structure approved assessment information. The assessment record remains human-led.
Behaviour information may contain patterns that staff want to understand.
Where enabled, AI may help organise selected behaviour data or highlight patterns for authorised human review.
That does not mean a pattern explains why something happened.
EdiWay AI should not:
Behaviour remains a professional, contextual and human-led workflow.
Safeguarding is a particularly important boundary.
EdiWay AI may assist authorised users with limited administrative or drafting activity where appropriately governed.
It must not replace the school's safeguarding process.
EdiWay AI does not:
Safeguarding concerns belong within the dedicated safeguarding workflow.
School leaders may need to understand information across many areas of the organisation. EdiWay AI can support appropriate leadership work around:
Ask questions about permitted school information.
Prepare narrative reports.
Organise key points.
Help structure appropriate school evidence.
Prepare working summaries around goals and actions.
Help prepare first drafts.
Organise relevant approved evidence.
Bring appropriate recorded information together. AI can help leaders navigate complexity. It does not certify that the school is compliant, inspection-ready or performing effectively.
Where enabled, EdiWay AI may assist authorised users with policy, compliance and evidence workflows.
Prepare a first working draft.
Help identify relevant permitted records.
Organise recorded information.
Help structure evidence already held by the school.
Draft possible follow-up for consideration.
Prepare summaries for leadership or governance discussion. The human and organisational boundary remains clear.
Where enabled, EdiWay AI may support appropriate workforce-development activity. That can include:
Suggest possible areas for professional development.
Help staff consider appropriate learning opportunities.
Prepare possible objectives for human agreement.
Help organise selected appraisal evidence.
Draft working next steps. AI assistance does not determine staff performance.
For consequential school workflows, the safest role for AI is usually to help create a draft. That creates a clearer sequence:
AI prepares a working output.
An authorised person checks it.
The user corrects or changes what is needed.
Important source information is checked.
The authorised human accepts the final content where the workflow requires it.
The reviewed output continues into the appropriate school process. Generating something is not the same as approving it. EdiWay’s governance direction is designed to preserve that distinction.
AI is more useful when users can understand the context behind an output. EdiWay's governance foundations are being built around source and provenance concepts such as:
Which school or learner information was used?
Which available source informed the output?
Which version of relevant information was available?
Which AI generation produced the draft?
What happened after generation?
Was important evidence unavailable or incomplete?
Can appropriate governance information about the generation be retained? Source-connected AI is an active area of product assurance. Not every EdiWay AI workflow should currently be assumed to expose a complete record-by-record source manifest.
AI behaviour can change when providers, models, prompts or policies change. EdiWay's AI-governance foundations are designed to treat those changes as something that needs control. That includes concepts such as:
Which AI service is being used?
Which model release produced the output?
Which governed instruction set applied?
What rules govern the workflow?
How is the behaviour reviewed?
Which configuration is active?
Can an inappropriate change be withdrawn?
AI-generated output may be incomplete, inaccurate or based on limited information. EdiWay's governance direction therefore includes making limitations visible rather than presenting model output as certainty. Users may need to consider:
Was relevant information unavailable?
Was some information outside the user’s permission scope?
Was the output based on only part of the relevant record?
Has the AI made an inference that needs checking?
Is the underlying information current?
Is there important context that exists outside the platform? AI confidence should never be treated as professional authority.
Asking AI a question should not create a route around information governance
Across EdiWay, AI should support professional work rather than quietly become the professional decision-maker. EdiWay AI does not make final decisions about:
The platform can help prepare information around these workflows. Authority remains with the authorised people and organisations responsible for the decision.
Use AI from the relevant teaching, learner, leadership or operational context.
Identify the authorised person making the request.
Keep school, learner, role and information permissions in view.
Use only appropriate information.
Prepare the requested working output.
Make missing context or evidence uncertainty clear where supported.
Check facts, sources, wording and professional context.
Record the relevant outcome or follow-up.
Keep consequential judgement outside the AI.
Retain relevant generation, revision or audit context where the workflow supports it.
EdiWay AI is the AI-assistance layer across the EdiWay education platform.
It is designed to support authorised users within their existing school workflows rather than operate as a separate unrestricted chatbot.
Natural-language assistance and data-search capabilities are represented within the EdiWay AI scope.
Access should remain permission-aware.
Yes.
Teacher-focused AI capabilities include lesson support, resource assistance, feedback support, reports and other productivity workflows.
AI report-generation capability is represented.
Generated reports should remain subject to appropriate human review.
Yes.
AI-assisted communication drafting is represented.
The authorised sender remains responsible for the final message.
Yes.
AI meeting-summary support is represented.
Yes, through appropriately governed drafting, evidence organisation and support suggestions.
It does not diagnose SEND or make statutory decisions.
AI may support preparation of working EHC-related drafts where enabled.
It does not issue, approve or amend a statutory EHC plan.
Intervention-support suggestions are represented.
Authorised professionals remain responsible for selecting, implementing and reviewing support.
No.
Safeguarding decisions remain with authorised safeguarding professionals.
No.
Behaviour and disciplinary decisions remain human.
AI can assist with feedback and marking-related workflows where enabled.
Formal academic judgements remain human-reviewed.
AI inspection-preparation support is represented.
It does not predict an inspection outcome or certify that the school is inspection-ready.
No.
AI may help organise information or identify areas for review.
Compliance determinations remain human and organisational responsibilities.
Permission-aware AI context is a core governance direction and has bounded implementation foundations.
Complete role, relationship, API and cross-school denial assurance remains part of ongoing testing.
They should not be treated as automatically approved records.
Consequential outputs are designed around human review, although complete receipt-bound write-back assurance across every workflow remains under development.
Model, prompt, policy, evaluation and rollback governance foundations are represented and continue to undergo release assurance.
No.
EdiWay AI is designed to reduce repetitive work and help people organise information while keeping professional judgement human-led.
Use governed assistance around teaching, resources and classroom work.
Support feedback, evidence review and reporting while keeping academic judgement human.
Use bounded AI assistance while preserving specialist professional authority.
Draft clearer communication while retaining human review and recipient controls.
Organise appropriate evidence without automating compliance decisions.
Understand the policies, controls, accountability and assurance behind EdiWay AI.
EdiWay is undergoing active development and live testing.
The current AI capability catalogue includes natural-language assistance, data search, report generation, parent communication drafting, staff support, SEND and EHC-related drafting assistance, intervention suggestions, meeting summaries, policy support, inspection preparation and compliance-support capabilities.
Teaching-specific foundations also include governed lesson support and AI-assisted feedback.
AI-governance foundations include permission-aware context, source/provenance controls, human review, model and prompt governance, and explicit limitation or evidence warnings. These controls remain under active assurance rather than constituting a blanket claim that every AI pathway is fully production-assured.
The current cross-assistant governance checkpoint also introduces durable generation, revision, model/prompt-release and governance-receipt foundations across a number of existing AI workflows, while complete source manifests, universal write-back controls and cross-role runtime evidence remain open.
EdiWay should therefore be positioned around governed, permission-aware and human-reviewed AI assistance—not autonomous education decision-making or unrestricted access to school data.