Sfg Research Solutions

04/03/2026

Admission is still Open
Call Now!

04/03/2026

What AI-Literate Students Do Differently!

I am revisiting a guide I published a few weeks ago on what AI-literate students do differently, and I want to highlight a few ideas from it that keep coming up in my conversations with teachers.

The guide outlines nine practices that separate students who use AI thoughtfully from those who use it passively.

A few that I keep returning to:

They use AI as a thinking partner. Students lay out their own reasoning first, then bring AI in to challenge, question, and offer feedback. The intellectual work stays with the student. AI responds to thinking that already exists.

They ideate themselves. The ideation process belongs to the student. Ideas built through course readings, class discussions, and personal learning experiences. AI can refine and push those ideas further, but the core thinking has to come from the student first.

They never copy and paste AI content. This one sounds obvious, but the research keeps confirming the cost. Overreliance on AI-generated content weakens cognitive engagement and long-term thinking abilities (Bai et al., 2024; Gerlich, 2025; Kosmyna et al., 2025).

They use AI for formative feedback. AI is available around the clock, it doesn't get tired, and it doesn't bring judgment into the interaction. But students need to be aware of sycophancy. AI tends to please, soften criticism, and reinforce what the student already believes. Prompts need to be explicit about asking for direct, critical feedback.

They reflect on how AI shapes their thinking. This is the metacognitive layer that makes everything else work. Students who pause to consider when AI helped, when it confused, and when it distracted them become more deliberate in how they use it.

The full guide is open access. It includes prompt examples for each of the nine practices.

Link in the first comment.

References
Bai, L., Liu, X., & Su, J. (2023). ChatGPT: The cognitive effects on learning and memory. Brain and Behavior, 13(10), e30.

Gerlich, M. (2025). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. Societies, 15(1), 6.

Kosmyna, N., et al. (2025). Your brain on ChatGPT: Accumulation of cognitive debt when using an AI assistant for essay writing task. arXiv preprint arXiv:2506.08872.

04/03/2026

What UNESCO's 2025 Report Means for Us Teachers and Educators!

I picked up UNESCO's 2025 volume on AI and education expecting confirmation of what I already think. Some of it was familiar. But a few arguments genuinely challenged me.

UNESCO's volume, AI and the Future of Education: Disruptions, Dilemmas and Directions, brings together 21 think pieces from scholars across philosophy, AI research, pedagogy, and policy.

One thread made me uncomfortable. Several contributors push back on the idea of AI as an efficiency tool. And I had to ask myself: when I encourage teachers to try AI, am I sometimes leading with efficiency? Save time. Generate rubrics faster. Produce feedback at scale.

I still think those are valid uses. But efficiency should never be the primary reason we do anything in education.

Education has purposes that go beyond speed: developing judgment, cultivating autonomy, building the capacity to think when the answer isn't obvious.

The equity argument is just as pressing. One-third of humanity remains offline. If AI-powered education becomes the default, who benefits and who gets excluded?

On assessment, the volume maps a tension I see every day. When we design around final products, AI becomes a shortcut. When we design around process and reflection, AI becomes a thinking partner.

My position hasn't changed. I still believe teachers should experiment with AI, critically and intentionally. But this report strengthened my conviction that the how matters enormously.

You can use AI in your teaching and still miss the point if you treat it as a plug-in for productivity.

The real work is building an AI pedagogy, and that project is far from finished.

Link in the first comment!

04/03/2026

In mid-2022, months before ChatGPT launched and the panic set in, Nakazawa, Udagawa, and Akabayashi published a quiet philosophical paper asking whether AI-assisted writing undermines the originality of academic researchers.

Their short answer was no. Their reasoning is what makes it worth reading in 2026.

The authors broke a research paper into its component parts and asked where AI involvement actually threatens originality.

Their argument: even when AI drafts text or generates candidate interpretations, the researcher still evaluates, selects, adjusts, and approves. And if AI eventually automates that evaluation step too, someone still has to evaluate the evaluation.

I've covered two other papers that converge on this same point from very different angles. Guetzkow, Lamont, and Mallard (2004) found that peer reviewers value "original approach," framing a problem in a new way, far above original results.

Johnson-Eilola and Selber (2007) argued that assemblage, building texts from existing materials to solve new problems, should count as legitimate scholarly composition.

Three papers, two decades apart, all saying the same thing: the lone-genius model of originality doesn't match how knowledge actually gets produced.

The paper's logic holds up. What hasn't held up is its optimism that researchers will always carefully evaluate AI output.

We've seen enough fabricated references and placeholder text to know that step gets skipped. Originality is protected only when researchers actually do the intellectual work. When they don't, it's not distributed originality. It's outsourced production.



References
Guetzkow, J., Lamont, M., & Mallard, G. (2004). What is originality in the humanities and the social sciences? American Sociological Review, 69(2), 190–212.

Johnson-Eilola, J., & Selber, S. A. (2007). Plagiarism, originality, assemblage. Computers and Composition, 24(4), 375–403.

Nakazawa, E., Udagawa, M., & Akabayashi, A. (2022). Does the use of AI to create academic research papers undermine researcher originality? AI, 3(3), 702–706.

19/02/2026

I think the plagiarism conversation in education is stuck.

We're still debating how to catch students, when the real question is how to teach them responsibility in a world where humans and machines write together.

Sarah Elaine Eaton (2023) makes this case in "Postplagiarism: Transdisciplinary Ethics and Integrity in the Age of Artificial Intelligence and Neurotechnology," published in the International Journal for Educational Integrity.

Her argument is clear. Human-AI co-authorship is already ordinary. When a student uses ChatGPT to brainstorm, draft, or restructure an argument, the resulting text is hybrid.

Tracing clean lines between human and machine contribution is practically impossible. And detection tools won't rescue us. OpenAI itself acknowledged their limits.

So what holds?

Responsibility.

"Humans can retain control over what they write, but they can also relinquish control to artificial intelligence tools if they choose. Although humans can relinquish control, they do not relinquish responsibility for what is written" (p. 5).

Eaton also reframes attribution as intellectual stewardship:

"Attribution is about knowing others' work, being able to speak to it accurately, and showing respect for others' contributions" (p. 6).

AI can generate citations. It can't participate in the relationships those citations represent.

We need to stop building integrity systems on surveillance and start building them on responsibility.

Link in the first comment.

Reference
Eaton, S. E. (2023). Postplagiarism: Transdisciplinary ethics and integrity in the age of artificial intelligence and neurotechnology. International Journal for Educational Integrity, 19(23).

18/02/2026

Over 200 AI Syllabi Policies From Educators Around the World!

In my last workshop with UAEU CETL and the AIGE Initiative, I stressed the importance of creating class-level AI policies.

These tend to work far better than broad institutional policies because they speak directly to the context of your course, your students, and your learning objectives.

I also talked about something I think deserves a lot more attention: AI disclosure as a teaching practice.

Make it a habit to disclose your own use of AI as a teacher. When you do, you model ethical AI use for your students in a way no policy document ever could.

And here's why I think ethical modelling is very important.

First, it's a live teaching moment. When you show students how you used AI to brainstorm discussion questions, draft a rubric, or organize your lecture notes, you're giving them a concrete example of what responsible AI use actually looks like. You also get to show them where the line is.

What counts as ethical use and what doesn't. These conversations are rich learning opportunities that build real AI literacy and help students make better decisions about how they use these tools.

Second, ethical modelling helps fight the stigma around AI use. Right now, a lot of students hide their AI use because they assume all of it is cheating. And that's a problem.

We need to show them that AI use exists on a spectrum. There are genuinely productive ways to leverage these tools to enhance learning, improve writing, and optimize workflows. But students won't believe that until they see their teachers doing it openly.

Once students see you modelling responsible use and engaging in honest conversations about it, something shifts.

The stigma starts to fade. They come forward, share how they've been using AI, show you their chat conversations, and ask real questions about where the boundaries are.

That's when your classroom moves from surveillance to shared understanding.

And that's where authentic learning happens.

Now, as I was preparing for the workshop, I came across an incredible resource put together by Lance Eaton, PhD.

It's a crowdsourced collection of AI syllabi policies from educators across disciplines, institutions, and countries. Hundreds of real examples showing how teachers are handling AI in their courses.

This is a treasure trove. You'll find everything from strict no-AI policies to full integration frameworks, along with creative approaches to citation, disclosure, and student reflection.

Use it as a starting point, adapt what fits your context, and build your own class-level policy from there.

I share the link in the first comment

18/02/2026

People Judge You When You Use AI (yes, I use AI). Here's the Research to Prove It.

I came across this paper while reading Ephrat Livni's article published in the New York Times. Livni reported on a New Zealand arson case where a judge discovered that the defendant's apology letters had been written with AI. He cut the sentence reduction in half.

The study Livni references is by Claessens, Veitch, and Everett (2026) titled "Negative Perceptions of Outsourcing to Artificial Intelligence".

The paper makes a very strong and valid argument. People judge you negatively when they learn you used AI.

They see you as lazier, less competent, and less trustworthy. And they judge your work as less meaningful and less authentic, even when the quality is identical to human-written text.

And not all tasks are equal.

Socio-relational tasks like writing apology letters, wedding vows, or love letters carry the heaviest penalty.

These tend to trigger deeper judgments about warmth, morality, and sincerity.

I agree with this completely. I would not want someone to use AI to apologize to me. Or to write me a eulogy. Some things need to come from a human being.

As for instrumental tasks like writing code, planning a schedule, or drafting a report, the penalty is milder. People seem to accept AI help when the task is practical and outcome-driven.

The paper is definitely worth reading.

Link in the first comment.

References
Claessens, S., Veitch, P., & Everett, J. A. C. (2026). Negative perceptions of outsourcing to artificial intelligence. Computers in Human Behavior, 177, 108894.

Livni, E. (2026, February 17). In arson case, a judge wrestles with A.I.-assisted apology letters. The New York Times.

18/02/2026

Can ChatGPT Help Students Learn Math?

To me, AI is probably one of the best things that ever happened to us in education.

But embracing AI and aggressively advocating for its use does not mean blanket use. We need guardrails in place if we want our students to benefit from its educational potential.

This study by Bastani et al. (2025) confirms this.

"Generative AI Without Guardrails Can Harm Learning: Evidence From High School Mathematics," published in PNAS, tested nearly 1,000 high school students across three conditions: no AI, open ChatGPT, and a guarded version designed with teacher input.

During practice, students with AI access did dramatically better. But when AI was removed for the exam, something troubling happened.

Students who had used open ChatGPT scored 17% worse than students who never had AI at all. They'd been asking for answers, not working through problems. "The vast majority of students are using GPT Base to obtain solutions" (p. 6).

The guarded version told a different story. It gave hints, not answers. It pushed students to try first. The harm disappeared.

And here's what concerns me most: students who performed worse "did not perceive that they performed worse or learned less" (p. 4). They had no idea.

AI without pedagogy behind it can hurt learning. And the damage is invisible to the very students it affects.

The fix isn't banning AI. It's designing how students interact with it. Guardrails that prompt thinking, resist answer-giving, and require engagement protect the learning process.

Link in the first comment!

Reference

Bastani, H., Bastani, O., Sungu, A., Geb, H., Kabakcı, Ö., & Marimane, R. (2025). Generative AI without guardrails can harm learning: Evidence from high school mathematics. Proceedings of the National Academy of Sciences, 122(26)

17/02/2026

I think UNESCO gave us the clearest blueprint yet for what AI literacy should look like in schools.
Their AI Competency Framework for Students (2024) covers four areas: human-centred mindset, ethics, AI techniques and applications, and system design.

The goal is to prepare "responsible and creative citizens that can co-create these desirable futures" (p. 12).

Learning progresses across three levels. Level 1 builds foundational literacy. Level 2 applies knowledge to complex tasks. Level 3 asks students to become AI co-creators, "developing human-centred solutions to positively impact the design and use of AI" (p. 21).

Most education systems are still at Level 0. That's the problem.

If you're already experimenting with AI in your classroom, this framework validates your work. If you haven't started, it gives you a structure. Either way, it's worth reading.

Reference
UNESCO. (2024). AI competency framework for students.

17/02/2026

I found this study through Matthew Connelly's opinion piece in the New York Times and thought it was worth revisiting.

Niloy et al. (2024) ran one of the first large experiments on ChatGPT and creative writing. "Is ChatGPT a Menace for Creative Writing Ability? An Experiment," published in the Journal of Computer Assisted Learning.

600 university students. Control group vs. experimental group with ChatGPT 3.5 access. Pre-test showed no difference between groups.

After using ChatGPT, the experimental group's creativity scores dropped significantly (p = 0.000). Accuracy and originality declined. "The generated content lost its originality compared to the state when ChatGPT did not intervene" (p. 926).

Two things did improve: elaboration and presentability. ChatGPT helped students expand ideas and structure content more clearly.

So the trade-off was specific. Clarity went up. Originality went down.

Now here's the thing. This was ChatGPT 3.5. Compared to GPT-5.2, 3.5 was primitive. No scaffolding was provided, no reflection, no pedagogical structure., and students were handed a tool and left alone with it.

We know from recent research that conditions matter. When students ask AI purposeful questions, writing improves (Cheng et al., 2025). When AI feedback is embedded in peer review and reflection, students evaluate it critically (Sperber et al., 2025).

The core insight still holds: passive AI use hurts creativity. That was true in 2024. It's true now. The tool has changed. The human tendency toward cognitive shortcuts hasn't.

Link in the first comment.

References
Cheng, Y., Fan, Y., Li, X., Chen, G., Gašević, D., & Swiecki, Z. (2025). Asking generative artificial intelligence the right questions improves writing performance. Computers and Education: Artificial Intelligence, 8, 100374.

Niloy, A. C., Akter, S., Sultana, N., Sultana, J., & Rahman, S. I. U. (2024). Is ChatGPT a menace for creative writing ability? An experiment. Journal of Computer Assisted Learning, 40(2), 919–930.

Sperber, L., MacArthur, M., Minnillo, S., Stillman, N., & Whithaus, C. (2025). Peer and AI Review + Reflection (PAIRR): A human-centered approach to formative assessment. Computers and Composition, 76, 102921.

06/10/2025

Enrol Now!!!

Al is the game changer!


AI
Digitalization

06/10/2025

https://techfocus24.com/why-every-ghanaian-school-must-teach-artificial-intelligence-and-why-you-should-enrol-now/

Artificial Intelligence (AI) is not the future. It is the now. Across the world, AI is reshaping how people live, work, and interact. From the phones in our pockets to the algorithms that determine what job opportunities, we see online, AI has become the invisible engine of modern life. The question