AI Literacy Assessment: 4 Workplace Skills Beyond Prompting

Prompting skills alone can no longer fully describe the AI capabilities employers need. Workday’s global workforce report, released on October 5, 2026, points to growing demand for more advanced, practical applications of AI.

An AI literacy assessment measures whether someone understands AI’s capabilities and limitations, can apply it to real work, can verify its outputs, and can take responsibility for the final decision.

Knowledge of tool names or prompting formulas provides only part of that picture.

This article helps HR and learning and development leaders identify what to include in an AI literacy assessment. It examines Workday’s findings alongside the assessment structure of Codepresso’s AI Fluent.

What Should an AI Literacy Assessment Measure?

An AI literacy assessment should establish whether employees can define a work objective, review AI-generated results, and turn them into a usable deliverable.

For workplace assessment, four areas deserve attention: AI understanding, practical application, output validation, and responsible judgment. Each can be evaluated through tasks relevant to the employee’s role.

Multiple-choice questions can help assess foundational knowledge across a large workforce. To understand practical capability, however, organizations also need evidence of how employees perform a task.

Two employees may produce similar answers through very different processes. Their instructions, error detection, and reasons for accepting or rejecting an output reveal important differences in capability.

Our Guide to Choosing an AI Competency Assessment Service explains how to evaluate assessment providers. Here, we apply those considerations to the four areas of AI literacy.

What Does Workday’s Report Reveal About Demand for AI Skills?

Workday’s report indicates that employers are increasingly seeking people who can build AI tools and automate workflows.

Demand for basic AI skills fell 25% after its January 2026 peak. Demand for AI tool development, workflow automation, and AI engineering grew 51% between September 2025 and July 2026.

These figures come from anonymized skills data in job postings from more than 550 employers using Workday Recruiting. They suggest that employers are becoming more specific about the capabilities they expect beyond experience using AI tools.

Workday also conducted a separate survey of 6,000 full-time employees and business leaders.

The 13-percentage-point gap between knowing which skills to develop and feeling supported by an employer points to a broader learning challenge. Offering a course is one step. Organizations also need to assess current capability, assign relevant practice, and evaluate progress using consistent criteria.

Is Prompt Training Enough for Corporate AI Education?

Prompt training is necessary, but it is only one part of working effectively with AI.

Employees also need to define tasks, validate outputs, improve results through revision, and apply their work in a real business setting.

Following a well-designed example during a workshop can produce a convincing result. Back at work, employees face additional requirements: internal document formats, customer data protection, approval processes, and department-specific quality standards.

Meeting those requirements depends on understanding the task and knowing how to review the result.

In the non-technical AI training examples discussed in our October 2 article, few programs explicitly described post-training assessment. Our subsequent guide to four steps for workplace adoption after generative AI training explained how task selection, deliverable creation, application records, and reassessment can form one connected learning process.

Prompting is the starting point. AI literacy also includes deciding whether an output is suitable for work—and knowing how to correct it when it is not.

A practical curriculum should therefore ask participants to select a real task, define what AI will do, identify what a person must review, and submit both the completed deliverable and a record of revisions.

How Do Workday’s Findings and AI Fluent Serve Different Purposes?

Workday’s report shows how demand for skills is changing across employers. AI Fluent assesses the current capabilities of people within a specific organization.

Codepresso’s approach uses role-specific tasks to evaluate both the working process and the final result against a shared competency model.

AI Fluent offers seven tracks and four competency levels: Entry, Beginner, Intermediate, and Professional.

Tasks vary by job function, while a shared scoring framework enables comparison across departments. Assessing prompt histories and intermediate outputs alongside final deliverables helps distinguish a strong result obtained by chance from one developed through deliberate review and revision.

For another example of connecting training with demonstrated performance, see Anthropic’s Practical Assessment and Reassessment After a 12-Week Residency.

As employers place greater emphasis on practical capability, internal training completion standards should also include evidence that employees can apply their learning at work.

How Does Individual AI Literacy Assessment Differ from Organizational Assessment?

An individual assessment identifies what one employee can do with AI and where they need development.

An organizational assessment reveals the distribution of capabilities across departments and highlights shared gaps that can inform training, role assignment, and governance.

Role-specific tasks combined with common scoring criteria provide a useful basis for these comparisons.

The purpose of organizational assessment is to identify common weaknesses and set development priorities.

One team may understand AI well but keep weak validation records. Another may be capable of automation but need stronger judgment about security risks.

Understanding these differences helps learning leaders direct resources toward the areas that need attention and tailor training accordingly.

How Should Organizations Measure AI Literacy Training Effectiveness?

Use the same competency model and practical tasks of comparable difficulty before and after training.

Completion rates and satisfaction scores remain useful operational indicators. Evidence of changes in working processes, output quality, and workplace application should be central to evaluating learning outcomes.

1. Assess before training. Use role-specific tasks to identify current competency and development needs.
2. Tailor learning to proficiency. Adjust task difficulty and learning content to assessment results, including differences among employees in the same role.
3. Require workplace deliverables. Ask participants to submit documents, analyses, or automation outputs that can be used in actual work.
4. Reassess using consistent criteria. Score prompt histories, intermediate revisions, and final results again to identify changes in performance.

This approach gives learning leaders more meaningful evidence for year-end reporting. Alongside the number of employees who completed training, they can explain which capabilities improved, in which roles, and where further development is needed.

Book a Demo →

Frequently Asked Questions