Every level of the organisation now needs some degree of AI literacy, not just technical teams. Alongside a genuine shortage of specialised AI talent, there's a growing need for critical thinking, AI oversight, prompt literacy and data literacy — skills that don't show up on a typical technical training plan.
Where this gets hard
- Significant AI literacy gaps exist across all levels of the organisation, not only among frontline technical staff.
- Specialised AI talent is genuinely scarce and expensive to hire or retain.
- Upskilling programmes are often aimed only at technical teams, leaving the majority of the workforce under-prepared.
- AI is automating routine tasks, forcing a redefinition of roles faster than job descriptions and career paths can keep up.
- Skills like critical thinking, AI oversight and data literacy are harder to train at scale than tool-specific skills.
Where to start
- Segment upskilling by role type — builders, users and overseers — rather than a single generic AI training programme.
- Prioritise critical thinking and AI-output evaluation skills as core curriculum, not an afterthought to tool training.
- Build career pathways that explicitly account for roles being redefined by automation, rather than leaving employees to guess.
- Use a mix of build, buy and partner approaches to close specialised AI talent gaps realistically.
- Measure upskilling success by demonstrated judgement and output quality, not attendance at a training session.
Download the consulting document for a role-based skills matrix and an upskilling roadmap template you can adapt to your organisation.
Does your upskilling plan reach the people who use AI daily, or only the people who build it?