Education technology products will have to meet safety standards under government plans announced by the Department for Education today.
Edtech companies would have to ensure products are designed to "minimise the risk of emotional or developmental harm to children", a DfE consultation states.
The announcement comes after a government minister warned of schools having to navigate a “Wild West” of edtech products.
'Heightened risks' from edtech innovation
The consultation says the scheme could "support safer decision-making, reduce burdens on education settings and give suppliers a clear and proportionate route to demonstrate that their products meet defined safety standards".
In a foreword to the consultation, DfE early education minister Paul Waugh said that "the pace of innovation and the growing complexity of products can present heightened risks".
Half of leaders said safeguarding and data concerns were to a great or some extent a barrier to increasing the uptake of technology by schools, in a Technology in Schools Survey published in 2025.
The certification scheme would set out the specified safety criteria an edtech product must meet to achieve certification, and the rules and procedures for assessing whether it meets those criteria.
It would not include an assessment on a product's effectiveness, as the DfE considers educational efficacy a "more nuanced judgement", it said.
AI tutors and SEND tools in scope
The products in scope for the certification scheme are filtering products, edtech conducting "high-risk data processing of children’s data", pupil-facing generative AI tools, and pupil-facing personalised learning tools.
High-risk data processing products include AI tutoring tools, special educational needs and disabilities (SEND) platforms and management information systems (MIS).
The move comes as the government develops its own AI tutoring tools, as Tes first revealed last year.
"When poorly designed and implemented, the use of generative AI tools can also present risks relating to children’s cognitive, emotional and social development and mental health, alongside intellectual property and data protection risks," the consultation said.
Minimum safety requirements
The consultation sets out minimum safety requirements for each of the four key product types it wants to certify.
Filtering systems would have to block access to harmful sites and inappropriate content "without unreasonably impacting teaching and learning or administration".
They will also need to have the capability to be applied to all devices managed by the education setting, including those taken off-site.
Tools like AI tutors, SEND platforms and MIS systems will have to conduct data protection impact assessments and comply with relevant intellectual property legislation and regulations.
They would also be expected to build upon the safety and data protection requirements within the DfE’s planned MIS Framework, the consultation added.
Major concerns have been raised about how the government plans to use data from the MIS framework, as Tes first reported last month.
Tes has asked the DfE if edtech products certified by the scheme would also have to contribute to the government's planned data spine, which would share information across education systems.
Pupil-facing tools, including AI, will have to comply with generative AI product standards, including mitigating any potential impact on user mental health and cognitive skills.
'Non-mandatory', for now
Under these proposals, using the product certification scheme would be "technically non-mandatory": the DfE does not have specific powers to require education settings to use certified edtech products or to require edtech suppliers to obtain safety certification for their products.
However, there would be a clear expectation to use it within statutory guidance.
The government said it would explore legislation in the future, as the absence of a legal requirement "may lead to reduced uptake and impact".
Ofsted could check if schools are using scheme
Ofsted could also be involved "through their consideration of safeguarding arrangements", the government consultation said.
The DfE is also exploring whether it will certify edtech products itself, or outsource this service.
Edtech products will have to recertify every three years according to the DfE's plans, with annual recertification for generative AI products.
To get essential briefings from Tes magazine delivered directly to your inbox, sign up to our newsletters. And stay up to date with the latest education news via the Tes magazine mobile app – available on iOS and on Android
