As part of our ‘staff and student perceptions of AI’ work, helping us understand current use and concerns around AI, I chaired two forums with Wales essential skills practitioners. For those not familiar with this, Essential Skills Wales covers communication, application of number and digital literacy, the core qualifications designed to develop learners’ reading and writing, practical maths and everyday digital capability for work and life.
These conversations also took place at a particularly important moment, as they were held alongside the national consultation on reforming Essential Skills Wales qualifications, which has now closed. The review is considering the future purpose, content and assessment of communication, application of number and digital literacy, with reformed qualifications expected to be introduced for teaching from 2028.
This wider reform context matters because the questions practitioners are raising about AI, assessment integrity and the risk of learners bypassing core skill development sit directly alongside the decisions Wales is now making about what ‘essential’ should mean in practice.
What came through most strongly was the now-familiar tension of increased AI use by staff and learners whilst working within qualifications that are still built around controlled authenticity, time-limited delivery, and high-stakes assessment rules.
Staff use of AI
Across both discussions, most practitioners described using AI in some form, even where learner use remains restricted.
As we see in other areas, the practitioners are using AI tools to help generate lesson materials quickly, tailor exercises to vocational contexts, draft handouts, summarise meetings, or support the administrative workload that sits behind teaching. Timesaving was particularly important to them as this is a particularly pressured sector.
Some practitioners also spoke about experimenting with more creative uses, such as generating audio resources for ESOL learners, creating interactive learning materials, or using AI tools to break down content into more accessible formats.
Again, as with other staff forums, we saw a strong understanding of the need to work with the AI tools, checking accuracy and appropriateness of outputs and so on.
Assessment
As expected, if one theme dominated both conversations, it was assessment.
Essential skills qualifications are often delivered in compressed timeframes, with limited contact hours and a strong focus on preparing learners for confirmatory tests and controlled tasks. In many contexts, AI is prohibited in assessment entirely.
We discussed the fact that this creates a structural contradiction in that staff may use AI to prepare teaching, but learners are expected to demonstrate competence without it. Practitioners described this as a difficult line to walk, especially when AI is already present in search engines, writing tools and everyday platforms.
Concerns were not only about deliberate cheating. Practitioners spoke about learners using AI without fully understanding the rules or being caught out by tools that automatically improve writing, such as advanced grammar assistants.
As for other sectors, there was a strong understanding that AI use is increasingly difficult to detect reliably, and AI detection tools are widely recognised as inaccurate. Formal writing styles, second-language writing patterns, and an employer’s ‘house style’ can all trigger false positives.
Instead, organisations are beginning to adopt clearer guidance frameworks, learner declarations, and assessment scale approaches that define acceptable use upfront. The emphasis is shifting from policing to clarity, transparency and shared understanding.
The erosion of essential skills
Practitioners described learners questioning why they need to learn writing, numeracy or problem solving when AI tools can generate answers instantly. Some learners already see these skills as unnecessary because ‘my phone can do it’. When answers are immediate, the process of working something out, struggling with it, making mistakes, and building understanding can start to feel optional.
In essential skills, where the goal is long-term capability rather than short-term completion, the risk that learning itself is bypassed is particularly high, especially as the tasks are often those easily replicated with AI.
We also discussed that these essential skills are the foundations that allow the learners to use AI critically – without basic maths and English they won’t have the tools or capability to evaluate AI outputs.
One participant also noted a practical issue: learners still need to do their job when devices fail or batteries die. Essential skills are what remain when the tool is no longer available.
Critical thinking
Although not always explicit within qualification specifications, critical thinking surfaced repeatedly as the skill that now underpins everything else.
Practitioners spoke about the growing challenge of helping learners identify reliable sources, extract meaning from long texts, and resist misinformation. Some are already using AI in structured ways to support this, for example, asking learners to summarise a document using AI and then explain why those points matter.
What is emerging is a shift in what ‘essential’ means, and part of this is critical thinking and its use in navigating a world which increasingly includes AI content that needs careful evaluation.
The difficulty is that essential skills delivery is already constrained. Time is limited and specifications are tight. Teaching often becomes teaching-to-assessment. The space for deeper critical development is frequently squeezed out.
Time constraints
Perhaps the most consistent constraint raised was time. Essential skills is often delivered as an additional requirement alongside vocational learning, with learners attending only short blocks of provision. Practitioners may not know learners well. Delivery models are built around efficiency and progression, not exploration.
Our forum attendees felt that in that environment, embedding AI meaningfully becomes almost impossible. Staff do not have the capacity to teach AI literacy, build reflective practice, explore ethical use, and still meet assessment demands.
The consensus was that this means AI cannot be separated from qualification reform. If AI is to be approached responsibly in essential skills, the system needs space to do so.
Conclusion
These discussions reveal a sector caught between two forces. On one side, essential skills education is built on authenticity, trust, and foundational competence. Those values matter deeply, and practitioners are right to protect them.
On the other side, AI is already embedded in everyday life and workplace practice. Learners will encounter it regardless of whether qualifications permit it.
So, the challenge is how essential skills education can evolve to help learners develop the judgement, literacy, numeracy and critical thinking needed to use AI effectively, without losing the foundational skills that underpin independence.
The hope from many is that qualification reform will recognise that AI is not an optional extra, but part of the environment learners are already living in, and that the system can respond whilst also avoiding negative impact on the existing core essential skills.
Find out more by visiting our Artificial Intelligence page to explore publications and resources, learn more about our communities and sign up for our AI Literacy training.
For regular updates from the team sign up to our mailing list.
Get in touch with the team directly at AI@jisc.ac.uk