Why ALiF asks this of organizations

When AI takes over the tasks through which people learn, expertise can fail to develop, wear away, or form around the wrong habits. Whether that happens depends largely on decisions that individuals do not control: which tasks are automated, whether learning happens in paid time, whether newcomers get supervised practice, and whether staff and trainees have a say. If individuals are expected to reach ALiF levels, the organizations that employ or train them share responsibility for the conditions that make those levels reachable.

What the Statement covers

An organization that implements ALiF for its own people and claims conformance would publish a Statement to the people in scope, with a named owner, a date and a version. It would set out:

1. Learning tasks. Which tasks in each role are learning opportunities and are, or may be, affected by AI, identified with the people responsible for education and supervision; and any roles whose tasks are proposed to be largely automated, with the development pathway offered to the people in them.
2. Practice and supervision. How protected practice, supervision and feedback are provided for trainees, new staff and staff without formal training structures; how trainees’ AI use grows with their level of supervision; the time and preparation supervisors get; and how staff in safety-critical roles keep their skills without AI.
3. Learning time. The time available, within working or study hours, for AI literacy development and protected practice.
4. Voice. A standing mechanism, which may be an existing body, with representation of trainees, staff (including administrative staff) and supervisors, that is consulted before AI changes their work, where concerns can be raised without detriment and responses are recorded.
5. Commitments behind decisions. Where ALiF levels are used in decisions about individuals, the commitments that make those levels attainable.

Where an element does not apply, the Statement says why.

Review and reporting

The Statement is reviewed at least every 12 months and before any AI deployment that adds, removes or changes a task it lists, through the organization’s existing approval process for AI. Each review reports what was actually delivered against the previous Statement and summarizes the concerns raised and the responses given.

What it does not do

It does not set minimum levels of provision. It asks organizations to state what they provide, deliver it, and report on it. Organizations may start with roles that have trainees or new staff and extend the Statement over time.

Timing

The requirement is proposed for ALiF 2.0. It would apply to conformance claims made 12 months or more after release. Organizations that claim conformance earlier would need a Statement by the end of the 18-month transition period. A template will be published with the Implementation Pathway Model v2.0.