Why GES’s 2026 National CPD Day Theme Could Change How Teachers Assess Learners

GES has chosen Assessment in the Age of AI as the theme for National CPD Day 2026. Here are 10 ways Ghanaian teachers can rethink assessment in the age of ChatGPT.

Article Information

Fact Check & Editorial Review

Status: Verified context, analytical article

The Ghana Education Service has declared Friday, September 25, 2026, as National Continuous Professional Development Day for teachers and staff, under the theme “Assessment in the Age of Artificial Intelligence (AI), Empowering Teachers to Assess Learning.” GES has directed Regional Directors to disseminate the information through the education management structure and coordinate arrangements for implementation.

This article goes beyond reporting the announcement. It examines what the theme could mean for classroom assessment and teacher practice.

Important distinction: The article does not argue that teachers should abandon traditional assessment. Rather, it argues that the widespread availability of generative AI requires teachers to reconsider which assessment methods provide the strongest evidence of actual learning.

Key Takeaways

1. AI has changed the assessment environment, even when schools have not formally adopted AI.
2. A learner’s final written product may no longer provide sufficient evidence of independent understanding.
3. Teachers should place greater emphasis on process, reasoning, application and demonstration of learning.
4. Oral questioning can become an important complement to written assignments.
5. Classroom-based and authentic assessment can provide stronger evidence of individual understanding.
6. AI detection tools should not be treated as infallible proof of academic misconduct.
7. Teachers need clear rules about when AI use is permitted, restricted or prohibited.
8. AI can be used to support teachers, but professional judgement must remain central.
9. Assessment should measure the intended learning outcomes, not merely a student’s ability to produce polished answers.
10. The GES 2026 CPD theme provides Ghana with an opportunity to develop a more sophisticated national conversation about assessment.

➡️ Related: GES Declares September 25, 2026 as National CPD Day

➡️ Related: AI in Education: What Ghanaian Teachers Need to Know

Introduction

Artificial intelligence has entered education faster than many assessment systems have been able to respond.

A teacher can now give a learner an assignment in the morning and receive a polished essay in the afternoon. The learner may have written every sentence independently, used AI for brainstorming, substantially edited an AI-generated draft or submitted an almost entirely machine-generated response.

The final document may look remarkably similar in each case.

This creates a fundamental assessment problem.

What exactly does the submitted work tell the teacher about what the learner has actually learned?

That question makes the Ghana Education Service’s 2026 National Continuous Professional Development Day theme particularly significant.

The Service has chosen:

“Assessment in the Age of Artificial Intelligence (AI), Empowering Teachers to Assess Learning.”

The theme is not simply about ChatGPT.

It is about a much deeper issue: how teachers can continue to obtain credible evidence of learning when learners have access to increasingly powerful tools that can produce academic work.

This is therefore a moment for Ghanaian teachers to rethink assessment, not to abandon it.

1. AI Has Changed the Assessment Problem

For decades, teachers have relied on assignments, projects, essays, take-home exercises, examinations and other forms of assessment to determine what learners know and can do.

These methods remain valuable.

The difficulty is that generative AI can now participate in the production of the evidence.

A learner can ask an AI system to:

• explain a concept;
• write an essay;
• generate revision questions;
• solve a mathematical problem;
• summarise a textbook chapter;
• produce computer code;
• translate a passage;
• create a presentation;
• improve grammar;
• generate ideas for a project; or
• provide feedback on a draft.

The educational question therefore changes.

Instead of asking only:

“Did the learner produce this answer?”

teachers increasingly need to ask:

“What evidence do I have that the learner understands this answer?”

That is a major shift in assessment thinking.

➡️ Related: ChatGPT in Education: Opportunities and Risks for Ghanaian Schools

2. The Final Product Is No Longer Always Enough

Consider a simple assignment:

“Explain three causes of environmental degradation in Ghana.”

A learner submits an excellent 800-word answer.

Under a traditional assessment model, the teacher may award marks based on accuracy, organisation, language and evidence.

But in an AI-enabled environment, another question becomes relevant:

Who produced the reasoning contained in the answer?

The teacher cannot necessarily determine that from the final product alone.

This does not mean that every impressive answer is AI-generated.

It means that the evidentiary value of the final product may have changed.

Teachers should therefore consider collecting evidence from different stages of the learning process.

3. Move From Product-Only Assessment to Process-and-Product Assessment

One of the most important changes teachers can make is to assess both the process and the final product.

For example, instead of assessing only the final essay, a teacher might assess:

Stage 1: Idea generation

The learner identifies the problem and proposes initial ideas.

Stage 2: Planning

The learner develops an outline or concept map.

Stage 3: Drafting

The learner produces an initial response.

Stage 4: Revision

The learner explains what was changed and why.

Stage 5: Final submission

The learner submits the completed work.

Stage 6: Reflection

The learner briefly explains what was learned from the task.

This creates a richer evidence trail.

A student who genuinely understands the subject should generally be able to discuss the reasoning behind the work.

➡️ Related: Authentic Assessment: Meaning, Types and Classroom Applications

4. Oral Assessment Could Become More Important

One of the simplest ways to establish understanding is also one of the oldest: ask the learner to explain the work.

Suppose a student submits an excellent research assignment.

Instead of immediately assuming that the student either wrote or copied it, the teacher could ask several brief questions:

• Why did you choose this argument?
• What evidence supports your conclusion?
• Which source did you find most useful?
• What would happen if one of your assumptions changed?
• Can you explain this concept without reading your paper?
• What part of the assignment did you find most difficult?

The purpose is not to interrogate learners.

It is to generate additional evidence of understanding.

This can be particularly valuable for project work, research assignments and presentations.

5. Assessment Should Measure the Intended Learning Outcome

AI also forces teachers to revisit an old but important assessment principle:

What exactly are you trying to measure?

Suppose the intended learning outcome is:

“Learners will be able to explain the causes of soil erosion.”

If the teacher gives a take-home essay, the task may measure several things simultaneously:

• subject knowledge;
• writing ability;
• access to digital resources;
• ability to use AI;
• ability to search for information;
• language proficiency; and
• ability to construct an argument.

But perhaps the intended outcome was simply understanding soil erosion.

The assessment task should therefore be aligned with the learning outcome.

AI makes this alignment even more important.

6. Teachers Should Not Fight AI With AI Detection Alone

A tempting response to generative AI is to search for a technological tool that can identify AI-generated writing.

This approach has limitations.

An AI detector should not be treated as a perfect judge of authorship.

False positives and false negatives are possible, and the reliability of detection can vary according to the tool, language, text characteristics and other factors.

A teacher should therefore avoid making a serious accusation of academic misconduct based solely on an automated detection score.

A stronger approach is to combine multiple forms of evidence.

For example:

Written work + classroom observation + oral explanation + drafts + learner reflection

provides a much richer basis for professional judgement.

The question should not simply be:

“Does this text look AI-generated?”

It should be:

“What evidence demonstrates this learner’s understanding?”

7. Teachers Need Clear AI Assessment Rules

Schools should not leave learners guessing about what constitutes acceptable AI use.

A teacher can establish different categories.

Category A: AI prohibited

AI cannot be used for the task.

This may be appropriate when the purpose is to measure independent performance.

Category B: AI permitted for support

Learners may use AI for brainstorming, explanations or feedback but must produce and understand the final work.

Category C: AI integrated into the task

AI use becomes part of the learning activity.

For example, learners might evaluate an AI-generated answer, identify its weaknesses and improve it.

This can actually turn AI into an object of critical thinking.

The key issue is transparency.

Learners should know what is permitted before assessment begins.

8. AI Can Become Part of the Assessment Itself

There is an important distinction between assessing despite AI and assessing through AI.

Teachers can design tasks in which students critically evaluate AI outputs.

For example:

“An AI system has generated the following explanation of photosynthesis. Identify five scientific inaccuracies or weaknesses and provide corrected explanations.”

This assessment can measure:

• subject knowledge;
• critical thinking;
• evaluation;
• scientific reasoning; and
• AI literacy.

In such a task, simply using AI is not cheating.

The educational objective is to determine whether the learner can evaluate AI intelligently.

This represents a potentially powerful direction for assessment in the AI era.

➡️ Related: AI Literacy in Schools: Why Students Must Learn to Evaluate AI

9. Local Context Should Become More Important

AI systems can produce generic answers very quickly.

Teachers can respond by designing assessment tasks that require learners to engage with their immediate environment.

For example:

Instead of asking:

“Discuss the causes of flooding.”

a teacher could ask:

“Identify two factors contributing to flooding in your community, provide evidence from observations made within the past month, interview one community member and propose two realistic interventions.”

Such a task requires local evidence.

The learner must observe, investigate, interpret and apply knowledge.

AI may help with some stages, but it cannot substitute completely for the learner’s engagement with the actual community.

This is especially important for Ghanaian classrooms.

Assessment should not become a competition between teachers and machines.

It should become an opportunity to connect learning with real contexts.

10. Practical Assessment Strategies for Ghanaian Teachers

Teachers do not need to redesign every assessment overnight.

Several manageable changes can make a significant difference.

Strategy 1: Use more classroom-based tasks

Allow learners to complete parts of assignments under teacher supervision.

Strategy 2: Ask learners to explain their answers

A short oral explanation can provide powerful evidence of understanding.

Strategy 3: Use progressive drafts

Collect evidence of how the learner developed the final response.

Strategy 4: Introduce reflection

Ask learners what they learned, what challenged them and how they solved the problem.

Strategy 5: Use unfamiliar scenarios

Require learners to apply knowledge to situations they have not previously encountered.

Strategy 6: Connect tasks to local contexts

Use community, school and Ghana-specific situations where appropriate.

Strategy 7: Assess reasoning

Do not focus only on the final answer. Ask learners to explain how they reached it.

Strategy 8: Use practical demonstrations

Where the subject permits, allow learners to demonstrate skills physically or through performance.

Strategy 9: Teach responsible AI use

Rather than pretending AI does not exist, help learners understand its opportunities, limitations and ethical implications.

Strategy 10: Combine assessment methods

No single assessment method can provide a complete picture of learning.

A combination of methods is often stronger.

11. What This Means for BECE and WASSCE Preparation

The rise of AI does not make examinations irrelevant.

In fact, it may increase the importance of assessment situations where learners must demonstrate independent performance.

Teachers preparing learners for national examinations should therefore continue to emphasise:

• conceptual understanding;
• problem solving;
• independent reasoning;
• application;
• interpretation;
• examination techniques;
• communication; and
• disciplined study.

AI should not become a substitute for learning.

A student who uses AI to generate an answer without understanding it may perform poorly when confronted with an unfamiliar examination question.

The fundamental educational objective remains unchanged:

Learners must know, understand and be able to use what they have learned.

12. What School Leaders Should Consider

The AI assessment conversation should not be left entirely to individual teachers.

School leaders have an important role to play.

Schools can begin developing clear policies around:

Acceptable AI use

What can students use AI for?

Disclosure

When should learners state that AI was used?

Assessment design

Which assignments should be completed under supervision?

Teacher capacity

What professional development do teachers need?

Data privacy

What learner information should never be entered into external AI systems?

Academic integrity

How should suspected misuse of AI be investigated?

A school-wide approach is more sustainable than every teacher developing completely different rules.

13. The Teacher’s Role Is Not Being Replaced

The arrival of AI does not eliminate the need for teachers.

It may actually make some dimensions of professional judgement more important.

AI can generate an answer.

The teacher determines whether the answer demonstrates the intended learning.

AI can suggest questions.

The teacher determines whether the questions are pedagogically appropriate.

AI can provide feedback.

The teacher determines whether the feedback is accurate, relevant and appropriate for the learner.

AI can produce information.

The teacher helps the learner determine whether the information is trustworthy.

The teacher therefore remains central to the interpretation of evidence of learning.

14. The Deeper Question: What Should We Assess?

The most important question emerging from the AI era is not:

“How do we stop students from using AI?”

It is:

“What kinds of human learning should education continue to value and assess?”

That question takes the discussion beyond technology.

Schools need learners who can:

• reason;
• question;
• evaluate evidence;
• communicate;
• collaborate;
• solve unfamiliar problems;
• make ethical decisions;
• apply knowledge;
• create;
• reflect; and
• exercise sound judgement.

These capacities become even more important when machines can generate fluent answers.

The challenge for teachers is therefore not to compete with AI at producing answers.

The challenge is to assess the quality of human learning behind those answers.

15. A Possible New Assessment Model for Ghana

The 2026 GES CPD theme could provide the basis for a broader Ghanaian assessment framework built around five complementary forms of evidence.

This model does not require teachers to abandon examinations.

Rather, it suggests that examinations should exist within a broader ecosystem of evidence.

For classroom assessment, the five dimensions could help teachers build stronger judgements about actual learning.

Frequently Asked Questions

Should teachers ban ChatGPT from all school assignments?

Not necessarily.

The appropriate approach depends on the learning objective. Some tasks may require independent work, while others may legitimately incorporate AI as a learning or evaluation tool.

Does using AI automatically mean a student has cheated?

No.

AI use can be appropriate or inappropriate depending on the instructions governing the assessment.

The critical issue is whether the use violates the assessment conditions or prevents the assessment from measuring its intended learning outcome.

Can AI replace classroom assessment?

No.

AI may support parts of assessment, but teachers remain responsible for professional judgement about learner performance.

Should schools use AI detectors?

They may be considered as one source of information, but an AI detector should not be treated as definitive evidence of misconduct.

What is authentic assessment?

Authentic assessment requires learners to apply knowledge and skills to meaningful, realistic or complex tasks rather than merely reproduce information.

Why is the GES 2026 CPD theme important?

It signals that assessment is becoming an important professional-development issue in Ghana’s AI-enabled educational environment. GES has officially selected AI and assessment as the focus of the September 25 CPD Day.

Conclusion: Ghana Should Not Wait for AI to Redesign Assessment

The Ghana Education Service’s 2026 National CPD Day theme arrives at an important moment.

Artificial intelligence has already changed the conditions under which teaching and learning take place.

The question is no longer whether Ghanaian learners will encounter AI.

They already do.

The more important question is whether Ghana’s assessment practices will evolve quickly enough to ensure that assessment continues to provide credible evidence of learning.

Teachers should not be expected to become AI engineers.

They should, however, become increasingly sophisticated designers and interpreters of assessment.

The future of assessment may therefore depend less on detecting machines and more on understanding learners.

A well-designed assessment should make it difficult for a learner to hide a lack of understanding behind a polished answer, while making it possible for a genuinely competent learner to demonstrate reasoning, application, creativity and judgement.

That is the deeper promise of the GES theme:

Assessment should not merely ask what answer a learner can produce. It should reveal what the learner can understand, explain, apply and defend.

And in the age of artificial intelligence, that distinction matters more than ever.

About the Author

Ferdinand Ellis is a Ghanaian education blogger, curriculum specialist, researcher and education policy analyst. He is the founder of EducateGhana.com, Ghana’s Education Intelligence Platform, and writes on curriculum, teacher education, classroom assessment, education policy and the responsible use of artificial intelligence in education.

Editorial Note

This article is an EducateGhana Education Intelligence analysis inspired by the Ghana Education Service’s declaration of September 25, 2026, as National Continuous Professional Development Day under the theme “Assessment in the Age of Artificial Intelligence (AI), Empowering Teachers to Assess Learning.”

The practical recommendations in this article are editorial analysis and should not be interpreted as official GES assessment regulations.

GES’s current public website identifies Circulars & Notices among its official resources, while current reporting independently confirms the September 25 CPD declaration.

Primary source: Ghana Education Service Headquarters, communication dated September 15, 2026, Ref. GES/DD-G(MS)/1/26/426.