Managing the risks and maximising the opportunities of AI for apprenticeships

Barely a day goes by without a prediction of how advances in artificial intelligence are going to (take your pick) destroy humanity, put us out of work, help us lead more fulfilling, productive lives or solve climate change.

A group of school leaders recently formed a body to advise and protect schools from the risks of AI, but in the world of apprenticeships there has mostly been silence. The DfE’s policy paper in March, titled Generative artificial intelligence in education, mentions research into the skills needed for future workforce training, but includes not a word on apprenticeships.

There is now a DfE call for evidence, which closes on 23 August, but in the wider economy things are moving much more quickly. AI is already being used for tasks previously done by entry-level staff. In legal services, AI tools can draft and summarise documents. Automated code generation, code analysis and debugging are encroaching on the work of software writers. Customer service roles are being replaced by AI-powered “bots.”

The jobs, apprenticeships, and T levels where change is most likely are at levels 2 to 5. In 2021/2, there were 306,000 apprenticeship starts at these levels. Those involving physical labour or where tasks are performed outdoors are at lower risk. All of which makes for an uncertain future, but there are things

providers can do now to prepare.

Experimentation as standard

All apprentices should use AI at least once. It doesn’t matter if it’s to prepare a first draft of research, to improve a document or email or to write some code, as long as it’s relevant to their job role.

This will allow them to safely explore the benefits and risks. AI is great at producing first drafts or suggesting amendments, but it gets things wrong. Have the apprentice check their work using authoritative sources. This leads to wider ethical conversations – about identifying fake news, plagiarism, how to spot AI “hallucinations” or problematic outputs.

In short, it’s an excellent primer on AI’s capabilities and on why we need to exercise caution and responsibility to use it productively.

Focus on what AI can’t do

Apprenticeships should complement strong foundational knowledge development with an increased focus on skills and behaviours where AI is poor, such as teamwork and higher-order thinking skills such as critical analysis, synthesis, evaluation, and creativity (as opposed to memorisation, repetition, and imitation).

Over two years ago, Damar partnered with O Shaped to embed its competency framework within the paralegal apprenticeship. The framework focuses on areas such as emotional intelligence and trust, problem solving and identifying opportunities. These are human-centric skills and behaviours where AI is unlikely to compete any time soon.

Use AI to widen access

Through Damar’s technology partnership with BARBRI  on solicitor apprenticeships, AI is adjusting the pace of learning to reflect individual strengths and weaknesses and help widen access to the legal profession.

This is just the start. AI can already create lesson plans for students at different levels and make sensible suggestions on syllabus sequencing. This gives us the chance to better personalise the learning journey, whether through recognition of prior learning, adjustments for SEND or other individual characteristics. We aren’t quite at the point of AI developing full, individually tailored training plans, but AI-assisted elements can already sit within a standard delivery model.

We can also use AI to improve marking and student feedback and help teaching staff manage their workloads and improve work/life balance. By asking students to use AI to evaluate their essays for ideas, content, sentence structure and organisation, we free up time for more detailed and valuable feedback and can spend longer on the highest impact teaching and coaching.

While some assessment methods will have to change, apprenticeships are well-placed to cope with concerns about “cheating”. Most end-point assessments feature a professional discussion where the assessor probes the apprentice’s understanding, so the apprentice cannot hide behind a nicely scripted AI presentation.

So, while we should be cautious, educators have less reason to be fearful than some may think. AI is here to stay, and apprenticeship providers who rise to the challenge will quickly see the benefits.