398: Google's Cheat Button Is Not Gone After All - Just Hidden

Plus, a new AI detection system in the UK wins a grant. Plus, another snippet from that Discovery Education report. Plus, International Quick Bites.

Issue 398

Subscribe below to join 4,899 (+3) other smart people who get “The Cheat Sheet.” New Issues every Tuesday and Thursday.

The Cheat Sheet is free. Although, patronage through paid subscriptions is what makes this newsletter possible. Individual subscriptions start at $8 a month ($80 annual), and institutional or corporate subscriptions are $250 a year. You can also support The Cheat Sheet by giving through Patreon.

---

More on the Google Cheat Button - It’s Not Gone, Just Hidden

After writing about the Google Cheat button — my term — and its coverage in the Washington Post (see Issue 397), a reader wrote me about it.

I asked if I could share some of their comments and their name. The reader agreed to sharing it, but I’m not sure I have clarity about saying who they are, so I’m not going to. Just to be safe.

In any case, this very credible reader says Google’s “homework help” button that was showing up during quizzes and exams and could answer on-screen questions in real time, is not gone, just hidden. And that students can still access it, if they want to.

The reader wrote:

I wanted to point out that while Chrome has removed the easy button in the address bar, there is still a way to access Google Lens in the browser. https://rtl.berkeley.edu/news/google-chrome-%E2%80%9Chomework-help%E2%80%9D-bcourses-quizzes

That link contains this information:

Update September 19, 2025: Google has removed the “Homework Help” feature from Canvas pages with the latest Chrome update. However, the functionality can still be accessed by manually launching “Search with Google Lens” from the Chrome settings menu or using the “Ask Google about this page” option in the browser URL bar.

So, that’s not great.

Our reader says more:

I [tested] in one of my sandbox canvas courses and pulled up a quiz as a student. I could go into Chrome settings and click Ask Google Lens for Help and voila! It sees I’m on a page with questions and will help me answer all of them quite easily. I’m removing Chrome from our machines and will be bringing this to the attention of our school’s Canvas team to hopefully make instructors aware of the vulnerabilities of using Chrome, even in the classroom.

I did see that Canvas Quiz logs do seem to track these interactions with Google lens as if the student ‘left canvas’, but many instructors don’t know about quiz logs and it would be a very tedious process to go through all the logs for a large class.

They also wrote subsequently that:

Instructors and admins should know the tool is still there, Google just hid the button. And it is pretty insidious. It could see that I was on a page with questions and was very ‘helpful’ in asking if I wanted help answering them...

Our friend also says that they use CMS software to block sites from being accessed during an assessment, but that the tool “isn’t able to block Google Lens when it’s opened from the settings menu.”

I hope that’s not too technical. And I do think it’s worth passing along, both as an alert for teachers and assessment supervisors, and as a request for Google, and the CMS providers, to do better.

Thank you to our reader who took the time to write.

New AI Detection Platform Wins $334,000 Grant for Testing and Deployment

From the UK, a new AI detection and analysis platform, developed at a university in Scotland, has won a grant to test and scale the system.

I’m not familiar with it, but I will ask for a demo and report back.

The news coverage includes:

Cyberhare Solutions, a University of the West of Scotland (UWS) venture, aims to help institutions spot when students use generative AI tools to produce essays, assignments, or research submissions without proper disclosure.

It comes amid a reported rise in AI-related cheating by more than 60% a year, which is costing universities an average of £150,000 a year to investigate.

Dr Stephen Langston, founder and CEO of the venture, and senior lecturer at UWS, believes the technology could save institutions millions while “restoring trust” in assessment.

First, I love when assessment security and academic integrity folks talk about restoring trust. It’s true, in my view. And it’s a nice dichotomy with the folks who say, and believe, that simply looking for cheating is a breach of trust.

I also had not seen the data on a 60% increase in AI-related cheating and the price tag. I’ll see if I can find that, but if anyone knows where it lives, please zip it over.

Also from the article, the system:

now being trialled across several other UK universities, IntegraGuard works differently from traditional “black box” plagiarism detectors.

Instead of flagging suspicious text, it provides evidence-based insights – from fake references to mapping out AI use in essays – and manages the whole referral process.

Mapping AI use in essays sounds like an AI detection system to me. Not sure how it’s different. But, as mentioned, I will see what I can find out. It’s interesting, for sure.

Also in the article a university leader said:

Academic integrity is a global challenge, and this work has the potential not only to protect standards in higher education, but to set a benchmark for how the sector responds to AI.

Imagine I put one of those “thumbs up” emojis here.

Another Snip from Discovery Education Report

In the last Issue, I wrote about the report from Discovery Education on student engagement — specifically, how 40% of K-12 students acknowledge using AI without permission.

I forgot to include this bit, also from that same report/survey:

70% Of students believe AI tools could help them complete schoolwork faster

Well, sure. Faster.

I’m still feeling good about the analogy (not mine) that AI use is academic steroids (see Issue 278). If all you cared about was going faster, why not?

International Quick Bites

• Headline in Fiji: “Exam cheating surge, major threat to Fiji’s education system”

• A headline from Ghana: “Exam malpractice ‘taking centre stage’ in Ghana — Ansa-Asare warns”

• In India, government minister says that, during another leaked national exam scandal, the “cheating jihadis” will be crushed.

• Also India: 23 people were caught cheating an Army test

• From Spain: “People cheat much more if they use AI: ‘It’s a level of deception we haven’t seen before’”

Class Notes:

• We picked up another generous paid subscriber, thank you ALG.

• Since one of the articles above is essentially from a reader, it’s a good place to mention that The Cheat Sheet is open to news, ideas, opinions related to academic integrity or fraud. To reach me, just hit “reply” to any Issue of The Cheat Sheet.

• Finally, if anyone in the Netherlands or EU would be willing to subscribe to this paper to get this article for me, I’d be very grateful. It’s about 15eu and you can cancel after, of course. I’m happy to reimburse you. I cannot subscribe due to bank/credit incompatibility with the US. Thank you!