#notebooklm really does have a hidden secret power.

I asked it to reimagine a research talk I first gave in 2006 on #contractcheating, and it produced this opening slide. There are ideas picked out in the slides that I'd long forgotten, and the approach makes the information seem fresh.

This is slide 3 out of 15 slides, all arguably much better than the ones I presented in 2006, when I first spoke about the contract cheating research that Robert Clarke and I had completed.

NotebookLM has extracted a key point from the paper and turned it into a clean visual story about how we identified and named contract cheating as a new threat to #academicintegrity.

One thing I like about this NotebookLM version is how it emphasises examples and case studies.

The map slide, for instance, isn’t one I could easily have made at the time (or would have thought to do so), but it tells the story clearly: contract cheating requests concentrated in English-speaking, Westernised countries, with some institutions much higher risk than others.

NotebookLM has also picked out key ideas from tables and the text, highlighting early trends we picked up on that matter now.

The profiles of high-volume users show this wasn’t just a few one-off cheats. It was already an organised industry, long before today’s gig platforms and #GenAI tools like #ChatGPT

There are obvious shortcomings in the NotebookLM version.

It doesn’t give the paper title, authors or full reference, and it defaults to US English, which just doesn’t feel right for a presentation on a UK paper. The slides are also not all that easy to edit, although I do have a workaround for that.

Even so, the core narrative of the 2006 research still comes through clearly.

What impressed me a lot to do was how NotebookLM didn't only process the paper text, but also analysed the images to work out the most important detail.

Here, it even extracted and recreated text from a low-quality screenshot in the original paper, recreated it, added useful labels and highlighting, and even kept one of the spelling mistakes in the redrawn slide (while quietly correcting another).

The contract cheating sites themselves have changed a bit since 2006, but many of the core ideas about this breach of academic integrity are still valid now.

To me, NotebookLM shows a clear opportunity to breathe new life into older academic work, and to bring forgotten details to the surface to share their importance with a new audience. That feels like a useful #GenAI aided direction for academic research communication.

Using NotebookLM to Rediscover the Hidden Details in Older Academic Integrity Research

This may be the bigger media plagiarism news story for a while. Apple TV+ has not only pulled a filmed TV series due to air shortly that had a 15 million euro budget, but also removed all marketing information and trailers for the show.

The Hunt was a six episode French language psychological thriller, scheduled to premiere in December 2025. The creator is accused of having adapted existing sources, without permission or acknowledgment, most notably a 1973 novel which had already been adapted into a film.

A further claim alleges that over 40% of the pilot episode was borrowed from other sources, including a graphic novel.

This is a complex situation for a show about a group of hunters who were being tracked by their rivals. Quite fitting as the plagiarism police appear to have been putting their own hunting expertise to good use here.

Scrubbed from Existence: How a Plagiarism Scandal Cost Apple TV+ Millions

I am proud to unveil my new course on how busy professors and instructors can save time and have exam scripts hand-marked for them using the latest AI technology.

(Yes, this Nano Banana Pro generated image is the early concept for the course cover)

#techethics #nanobanana #academicintegrity

I haven’t generated the full sales copy image yet, so I don’t know what to charge.

But if this really hands exam scripts over to an AI marker… it must be worth a lot, right?

What would you pay to learn how AI can mark handwritten exam papers for you? #academicoutsourcing

I’ve already generated most of the teaching materials. Just in case anyone wants to hire me.

It’s fun once you get going. The central process diagram #nanobanana generated is quite beautiful. Really, it all feels like high-value material. I wouldn’t feel comfortable selling this to the educational market, but I would sell to other markets using this material generation approach.

More seriously, AI exam script marking is possible now using Gemini Nano Banana Pro, and it’s not difficult to set up.

I’d need to put in time to learn the fine details and best approaches, but I personally don’t plan to go down that route. I’m sure others already have.

Next-Level AI Grading: How Much Is Your Weekend Worth?

I’ve been finding my guru branding on X just isn’t strong enough any more, so I’ve worked with Nano Banana Pro to produce this concept image of what my new social media profiles could look like.

What do you think?

If nothing else, I’m impressed by how fast image generation technology has come on, and how aligned the marketing message and visual presentation on the profile is.

I won’t say making this image was foolproof or one-click (yet). I did go through multiple rounds of edits, feeding in lots of ideas and corrections. I also redid the process multiple times to reduce the edit artefacts that appear.

Even then, this version isn’t perfect, but imagine now how easy it is now for those in the know to both makeover profiles “for good” (it seems fitting to quote Elphaba this weekend), but also to mislead.

I’m not rebranding just yet. But this gives me an idea for a new self-updating social media strand, DrLancasterAI.

#techethics #nanobanana #academicintegrity #GenAI #DrLancasterAI

I Asked an Image Model to Turn Me into a Thought Leader. It Understood the Assignment.

Do you remember Cluely, the website that launched earlier this year advertised at helping students to cheat using AI in online exams, then raised $15 million? Cluely is now the number 1 "AI assistant for meetings", that takes notes for you and displays them on an invisible overlay. Aren't there a lot of these already? Maybe this time, investing in cheating doesn't pay?  #academicintegrity

Cheating Doesn't Pay

Move over Wikipedia! With the launch of Grokipedia, another source of truth is coming for student essays and written assessments. And, with it, all kinds of #techethics implications.

Grokipedia is the brand new Wikipedia competitor, with all the entries generated using the Grok AI tool. I've looked at them. They're detailed and full of references.

Now, Grokipedia is not perfect. Entries are not human checked, and the references are often not to particularly great sources, but they are ones easily available for GenAI tools to crawl. There may also be political biases.

But, what Grokipedia does offer is the equivalent of over 800,000 deep research articles, all in a consistent style. Useful in their own right, but there are all kinds of creative ways these could be used, including for training small models for local needs. Yes, AI training AI.

This could also be the first of many GenAI written online encyclopaedias. It takes financial investment to generate 800,000 articles, but it's not a technically complex task. Imagine all the different viewpoints that could be presented, and which could feature in student work.

Is the next GenAI age assessment going to be to require students to check Grokipedia, fact check a page, and provide independent verification? Or should I instead be the first person to claim the phrase "AI-encyclopaedia recycling" to describe an #academicintegrity breach?

Grokipedia and the Rise of the AI-Encyclopaedia: What It Means for Assessment

"Anecdotally, British professors in the ’80s sometimes received suspiciously well-written papers and suspected third-party help, but without services like Turnitin (which did not emerge until 2000s), detection was difficult unless the student admitted purchase."

Anecdotally, in this instance, means hallucinated by ChatGPT - which is a different way of saying ChatGPT has made something up that sounds convincing and may be true, but which ChatGPT could never have been told.

It isn't the hallucination that fascinates me, more that ChatGPT has chosen to use the word "anecdotally" here to disguise it.

#academicintegrity #contractcheating

An Anecdote From ChatGPT

With the International Day of Action for Academic Integrity now complete (for 2025) what were my 5 top take-home messages from the 15 hours of live presentations? I've compiled this with help of Mary Davis' summary as I didn't quite manage to attend all 15 hours myself. #IDoA2025 #academicintegrity

Message 1: Student voices surrounding academic integrity issues are powerful. We heard real life examples of resilience and lived realities in warzones going far beyond the abstract situations often used. We need to reconsider the way we engage with and train staff.

Message 2: It really is hard to be honest all of the time. This came through clearly from the gathering of psychologists, who stressed how moral disengagement, pressures, and the pro-social agency shape student views beyond education.

Message 3: Creativity engages students in ways that academic integrity policies can't. Playful and participatory formats, like the videos, costumes, superpowers, and even thinking like Dolly Parton, help generate a sense of belonging and ownership beyond formal messaging.

Message 4: Integrity compounds over time. A small choice today can be a pivotal moment for learning. There's a long-term cost to many decisions that can shape careers and the culture in organisations. Misconduct can damage livelihoods and important to discuss this with students.

Message 5: We need to extend the academic integrity movement. Many presenters shared ideas, resources and interactive tools. Mary closed the event by quoting students and reminding us to feel empowered, and to be kind to ourselves on the long journey.

Do watch the videos when the International Center for Academic Integrity team make them available. There are even discussions of AI. Now, what will the IDOA theme for 2026 be? #IDoA2026

Be Kind, Be Honest, Be Creative: Five Practical Lessons from IDoA 2025

The International Day of Action for Academic Integrity is in full swing. Earlier we met the superhero "Role Model Radiance" (known in her unmasked identity as Professor Mary Davis) and heard about excellent student-led activity taking place around the world.

The panel considered ethical AI use in education raised a lot of concerns surrounding global injustice, hidden biases, exploitative development models, environmental impact, and moral damage. There were also academic concerns about eroding student skills and confidence, and considering how assessments are designed and integrity framed. Lots to think about, but this is a reminder of how GenAI isn't automatically accepted across education.

#academicintegrity #IDoA2025

The Hidden Costs of Generative AI - Environmental, Ethical and Pedagogical

I had a very wide ranging and enjoyable conversation with Irene Glendinning tonight about #academicintegrity and #contractcheating in the run up to the International Day of Action on October 15 2025. I hope that the video will be available shortly.

We discussed Irene's well-deserved lifetime achievement award for her academic integrity work and how her retirement is proving to be anything but quiet.

We had a long chat about The Shadow Scholars documentary, the filming timeline, and how AI had impacted both the story itself and the visual presentation of the participants, as well as the ethical complexities surrounding the contract cheating industry.

As Irene reiterated, the contract cheating industry has not gone away, only changed how it operates in light of AI. I even discovered some new things about the industry. I had no idea one essay mill used to give free package holidays away to students to enjoy while they completed their assignments.

This all barely scratches the surface of our conversation, so thanks to everyone who joined us and posted in the chat. Roll on the main International Day of Action where 15 hours of academic integrity work from around will be showcased. https://academicintegrity.org/aws/ICAI/pt/sp/idoa