University cancels publication of coding competition results over AI cheating fears

[midian182]

A hot potato: Once again, the problem of AI being used by students to cheat on exams has reared its head. On this occasion, the University of Waterloo's Canadian Computing Competition (CCC) has decided not to publish its official results, which it usually does, over the belief that AI was used by some participants to write code.

Those who do well in the University of Waterloo's CCC are often accepted into the University's prestigious computing and engineering programs, or are even selected to represent Canada in international competitions, student Juan Marulanda De Los Rios told The Logic. It can also help when applying for internships, jobs, or work experience programs.

The University normally releases students' CCC scores every year, but co-chairs J.P. Pretti and Troy Vasiga said in a statement that the 2025 results won't be made public. The reason for this decision is that it is "clear" many students submitted code they did not write themselves, relying instead on "forbidden external help."

The co-chairs add that because of the cheating, the reliability of ranking students would not be equitable, fair, or accurate.

Those forbidden external tools include the use of AI. But it seems some students didn't adhere to these rules. University of Waterloo spokesperson David George-Cosh declined to comment about how many people cheated during the competition or which AI tools were used.

Teachers usually supervise screens during the test as a way of preventing cheating, but one teacher sometimes has to monitor several students at once, and there aren't strict barriers to bringing code into the competition or a way of restricting access to websites and applications, said one person. And thanks to the integration of Copilot into GitHub, students are now able to cheat using AI without closing their programs.

Other coding competitions are struggling to prevent generative AI from being used, especially those that allow participants to access different websites or even take the competition home.

The University of Waterloo says it will introduce additional measures to safeguard future competitions, which includes improved technology, supervision, and clearer communications for students and teachers.

Students using AI to cheat isn't something new – the problem was described as endemic last August. There have also been instances where parents have sued schools who punished their children for using AI in exams.

Permalink to story:

University cancels publication of coding competition results over AI cheating fears

 

[Baldrickalpha1]

Pretty simple fix, no phones, no Internet, performing tests on preset machines that don't have copilot or any other ai installed.
Any of those discovered, instant disqualification, no appeal.

 

[gamerk2]

Pretty simple fix, no phones, no Internet, performing tests on preset machines that don't have copilot or any other ai installed.
Any of those discovered, instant disqualification, no appeal.

The problem is that costs organizers money, but yes, preset machines with no external connections are the simplest solution.

That being said...I'm a professional and absolutely am starting to roll AI into my coding workflow. Which begs the question what is being tested here: Knowledge of the language, or program design? Because AI is much worse at the latter.

 

[VitalyT]

It has been known for over 100 years, that the weakest link in any automation is the human being, to become ultimately redundant. And they made scary movies about that long ago, about such a doom day in the future. And now that it is here, this generation of imbeciles is embracing it. I say we all deserve to f-ng die off, like dinosaurs, 'cos it's coming anyhow.

 

[Squid Surprise]

The obvious solution is to allow AI - and change the contest structure so that you “grade” the human element.

AI isn’t going away, and banning it basically makes your contest irrelevant.

It would be like making a math contest only arithmetic - calculators have existed for decades and math contests evolved to allow them yet still be relevant.

People always whine about new technology - how about USING that new technology instead of trying to pretend it doesn’t exist?

 

[ScottSoapbox]

thanks to the integration of Copilot into GitHub, students are now able to cheat using AI without closing their programs

Why is this still cheating? At the point an AI assistant is integrated into the program used to write the code (and is standard practice in the industry), simply allow everyone to use it.

This reminds me of a computer science class where we had to write our code for exams with pencil and paper and then lost points for every spacing or comma error because, you know, that's what you're paid for: coding grammar, not solving complex problems with elegant code.

Allow AI and make the coding problems harder. Problem solved.

 

[RudyBob]

Cheating is a way of life for many

 

[Theinsanegamer]

The obvious solution is to allow AI - and change the contest structure so that you “grade” the human element.

AI isn’t going away, and banning it basically makes your contest irrelevant.

It would be like making a math contest only arithmetic - calculators have existed for decades and math contests evolved to allow them yet still be relevant.

People always whine about new technology - how about USING that new technology instead of trying to pretend it doesn’t exist?

The education system, despite it's stated purpose, is almost always decades behind the real world. Giving up control and allowing innovative new learning techniques or tools is the LAST thing they want.

I remember being told we were not allowed to use calculators in math class because "your not always going to have a calculator in your pocket". This was half a decade after the iPhone released.

 

[NotYourITguy]

If you aren't using AI and are a developer today, you're woefully behind. This university like all others should figure out real quick how to make use of modern tools instead of crying wolf about how people are "cheating". The reality is employers and founders can use whatever they can afford to get their hands on.

It's frankly a waste of time to have a contest that prevents people from reaching their full potential. Let people come up with the best solutions they can using all available resources.

 

[New Day]

I remember being told we were not allowed to use calculators in math class because "your not always going to have a calculator in your pocket". This was half a decade after the iPhone released.

What? College algebra teachers in the 90s required us to have graphing calculators.....

 

[magoid]

Some people seems really confused here. You would not allow calculators on simple math tests where the teacher is checking if your learned basic, simple math stuff, not advanced courses where it would take you half a day of calculations if you doing it by hand.

You still need to know math to use a calculator, especially for advanced stuff. The problem with AI is that it can do the whole work for you, no basic understanding of coding needed. Is like copying all the answers of the kid next to you in a written exam.

There is a difference in using AI in day to day for repetitive tasks, and learning real coding skills at school. You go to school to learn how to code, not to ask AI to do the whole thing for you. If you really think AI will save your *** when the real problems arrive, think again.

 

[Squid Surprise]

There is a difference in using AI in day to day for repetitive tasks, and learning real coding skills at school. You go to school to learn how to code, not to ask AI to do the whole thing for you. If you really think AI will save your *** when the real problems arrive, think again.

It doesn’t matter… education facilities need to understand that AI is here to stay - and everyone will be using it in the future (and most are using it now). Simply banning it is 1) unenforceable and 2) not “educational “.
Educational facilities are SUPPOSED to be teaching students how to function in the real world - and that world will be using AI.
We should be being taught how to use AI in tasks - and contests need to change based on current technology.

Decades ago, I remember math classes being worksheets where we mindlessly solved arithmetic questions…. While it was partially useful to be able to do this, it ignored the fact that other than very basic arithmetic, we didn’t really need to be able to do this…
SLOWLY, most schools have adapted (some quite poorly alas) by changing math curriculums into learning “why” 2+2=4 instead of just mindlessly solving it… it has taken SO long for schools to do this… and most are still doing it poorly…
AI is here - while I’m sure it will take schools decades to adapt to teaching with it, they SHOULD be doing it now - I’m sure good teachers already are (but their administrators won’t understand for years to come).