Artificial intelligence is becoming part of everyday education.
Students use generative AI to research, explain concepts, draft content and solve problems.
Teachers use it to prepare material, personalize learning, reduce administrative work and explore new ways of supporting students.
Educational institutions are beginning to integrate AI into digital learning environments, productivity platforms and administrative processes.
The question is no longer whether AI will enter education.
It already has.
The more difficult question is:
Can institutions understand, govern and measure AI use quickly enough to ensure that it actually improves education?
Two reports published by the European Commission in September 2026 highlight this challenge.
The Commission describes generative AI as moving steadily from technological innovation toward a mainstream tool in European education. At the same time, it stresses that the effect of generative AI on learning depends heavily on how the technology is used, while its longer-term impact on skills remains uncertain.
That creates an important distinction.
AI adoption is not the same as AI readiness.
Generative AI Is Moving Into the Educational Mainstream
The European Commission’s paper Friend or foe? Evidence from the use of generative artificial intelligence in learning and teaching describes a technology that is increasingly becoming part of normal learning and teaching practices.
The potential benefits are significant.
Generative AI can support personalized learning.
It can provide additional assistance to students with different learning needs.
It can help teachers create educational material.
It can reduce some administrative workload.
It can give learners access to explanations, examples and support at a scale that would previously have been difficult to provide.
But the same technology can also create new risks.
The Commission highlights issues such as academic integrity, cognitive overreliance, privacy and ethical concerns.
Most importantly, the research indicates that the outcome depends on how AI is used.
When students use generative AI to augment their own knowledge and actively engage with the material, the technology can support stronger learning outcomes.
When AI-generated information is consumed mechanically and accepted without deeper understanding, the effect can be very different.
That distinction has major implications for AI governance.
The relevant question is not simply:
Do our students and teachers use AI?
It is:
How is AI being used, under what conditions, and toward what educational outcome?
AI Literacy Is Becoming Part of AI Governance
Governance in education cannot be separated from capability.
A policy can tell a teacher which AI services are permitted.
It cannot automatically teach that person how to evaluate an AI-generated answer.
A school can publish responsible-use guidance.
That does not automatically give students the ability to recognise hallucinations, bias or weak reasoning.
An institution can purchase an enterprise AI platform.
That does not guarantee meaningful or responsible use.
This is why the Commission emphasizes substantial initial and continuous professional development for teachers.
AI literacy increasingly needs to include more than the ability to operate an AI tool.
It should help people understand questions such as:
What can this AI system do?
Where can it be wrong?
What information should not be entered?
When should its output be verified?
When should human judgement take priority?
What does responsible use look like for this particular learning activity?
In that sense, AI literacy becomes part of the control environment.
Governance defines the boundaries.
AI literacy helps people operate safely and effectively within them.
Policy Is Growing, But Implementation Still Matters
The European Commission’s second report, Digital education at school in Europe 2026: Bridging gaps in access, teaching and learning, shows that European education systems are actively responding to digital transformation.
Digital competence is already part of lower-secondary curricula across almost all systems analysed.
Nearly two-thirds of education systems have also developed strategies, guidance or other policy frameworks addressing AI.
That is substantial progress.
But having a framework and operating that framework are different things.
Implementation conditions still vary across education systems, including access to infrastructure, technical support and digital environments.
AI adds another layer of complexity.
An educational institution may formally approve one set of technologies while students and staff discover many others independently.
A teacher may use an AI assistant through a personal account.
A student may upload coursework to an external generative AI service.
A department may subscribe to a specialist AI application.
An existing education platform may introduce a new embedded AI capability.
A developer or research team may access models through APIs.
The institution may therefore have AI policies while still lacking complete visibility into the AI actually being used.
This is where education begins to encounter the same problem already appearing across other industries:
Shadow AI.
Shadow AI Is Also an Education Problem
Shadow AI does not require malicious intent.
In education, it may simply begin with a student, teacher or administrator trying to solve a problem more efficiently.
But the governance implications can be significant.
Consider a student using an external AI service to review an assignment.
What information was uploaded?
Was personally identifiable information included?
Was the account institutional or personal?
How will the provider use or retain the data?
Does the institution know that the service is being used?
Now consider a teacher using AI to analyse student work.
The same questions become more consequential.
Student records, performance information, behavioural information or other sensitive educational data may be involved.
The issue therefore extends beyond whether an AI provider has been approved.
Institutions increasingly need to understand:
which AI services exist,
who is using them,
which accounts are involved,
what information they can access,
what purpose they serve,
and whether their use aligns with institutional policy.
Without visibility, responsible-use guidance can only reach the AI activity the institution already knows about.
Privacy Cannot Be Separated From AI Adoption
The European Commission specifically identifies privacy among the risks that education stakeholders need to address.
That is particularly important because education environments contain sensitive information about students, teachers and families.
AI use can potentially involve:
student identities,
academic performance,
learning difficulties,
behavioural information,
assignments,
teacher feedback,
administrative records,
images,
and other personal information.
The governance question is therefore not merely whether an AI tool produces a useful answer.
Institutions need to understand what information reaches the AI system and under what conditions.
That means responsible education AI increasingly intersects with:
identity,
data protection,
procurement,
security,
policy,
ownership,
and auditability.
This also illustrates why AI governance cannot remain only a document.
A policy may prohibit certain data from being entered into external AI systems.
Operational governance requires enough evidence and organizational visibility to understand whether the rule is actually being followed.
Education Needs Visibility Before It Can Have Effective Governance
Institutions cannot meaningfully govern an AI environment they cannot describe.
A useful starting point is therefore an AI inventory.
But that inventory increasingly needs to extend beyond a list of purchased applications.
Educational organizations may need visibility across several dimensions.
Providers and models: Which AI services and models are being used?
Applications: Which learning, productivity or administrative applications contain AI capabilities?
Users: Which students, educators, administrators or teams use them?
Ownership: Which department or organizational function is responsible?
Approval: Is the activity institutionally approved?
Data exposure: What information may be processed?
Usage: How extensively is the capability being used?
Governance: Which institutional requirements apply?
Outcome: What educational or operational result is the AI expected to support?
This turns AI inventory into something more useful.
It becomes management context.
Responsible AI Use Also Requires Evidence of Outcomes
The Commission’s findings raise another important issue.
AI can be widely used without necessarily improving education.
A school may have thousands of AI interactions.
A university may provide AI access to every student.
Teachers may adopt AI across multiple subjects.
None of those metrics proves that learning has improved.
Usage measures activity.
It does not automatically measure outcome.
For educational AI, organizations increasingly need to connect:
AI Usage → Educational Activity → Accepted Output → Learning or Operational Outcome → Measurable Value
The outcome will differ depending on the use case.
It might involve:
better understanding of a subject,
improved learning support,
reduced administrative workload,
faster preparation of teaching material,
greater accessibility,
higher-quality feedback,
or another explicitly defined educational result.
The Commission’s research makes this distinction especially relevant because the impact of generative AI depends on how people engage with it.
That means one of the most important questions for education is not:
How much AI are we using?
It is:
Is this AI use helping us achieve the outcome we intended?
From Education AI Policy to an AI Management Layer
This is where AssetUno AI approaches the issue from a broader enterprise management perspective.
The goal is not to replace learning platforms, security controls, identity systems or institutional education policies.
It is to connect available evidence from different parts of the AI environment into a management context.
For an educational organization, that can involve several connected capabilities.
AI Discovery
Establish which AI providers, applications, models and services exist across the organization where evidence is available.
Shadow AI
Identify unmanaged or insufficiently governed AI activity that may require investigation or additional organizational context.
Organizational Context
Connect AI evidence with users, teams, departments, projects, cost centers or responsible owners.
Usage & Cost
Understand adoption and available consumption or financial evidence across AI services.
Governance & Risk
Connect AI activity with ownership, policies, risk evidence, exceptions and management decisions.
AI Value & Outcomes
Move beyond AI consumption by connecting activity with accepted work and measurable educational or operational outcomes.
The objective is not simply to create another technology inventory.
It is to make institutional AI activity visible, attributable, governable and measurable.
The Next Phase of AI in Education Is Responsible Management
The European Commission’s 2026 findings show an education environment that is already moving beyond the question of whether generative AI should exist.
It does exist.
It is being used.
Policies are developing.
Teachers and students are adapting.
The next challenge is institutional maturity.
Educational organizations increasingly need to answer:
What AI is being used?
Who is using it?
Is it institutionally approved?
What information does it process?
Do users understand its limitations?
Which policies and responsibilities apply?
Can the institution identify unmanaged use?
Is the AI supporting the educational purpose for which it was introduced?
Can that outcome be demonstrated?
The future of AI in education should not be measured by how quickly institutions can introduce more AI.
It should be measured by whether they can use AI responsibly, deliberately and with evidence that it improves the outcomes that matter.
That requires more than adoption.
It requires visibility.
It requires AI literacy.
It requires governance.
And it requires the ability to connect technology use with educational value.
Primary Reference: European Commission, Two Commission reports show impact of artificial intelligence and digital technologies on teaching and learning in Europe, 22 September 2026
The European Commission article presents findings from two 2026 reports examining generative AI in learning and teaching and the broader state of digital education across European education systems.
