What Chess Teaches About Academic Integrity

In 1770, Wolfgang von Kempelen unveiled the Mechanical Turk, a revolutionary chess automaton that was purported to outmatch any human opponent.

Despite the impossibility of such a feat in the 18th century, the “machine” was a resounding success. It managed to play and defeat notable opponents such as Napoleon Bonaparte and Benjamin Franklin, even though it was, in reality, just a human crammed into a tiny box.

Despite the suspicions that the Turk was a fraud, it managed to complete its eighty-year run without being publicly debunked. It was only afterward that the truth about the “invention” came out.

Today, the bots are significantly more intelligent, and chess has the opposite problem: bots pretending to be humans. This has led to countless allegations of players cheating, including some very outlandish claims.

Though chess is more popular than ever, it’s also struggling to maintain its legitimacy. When a free, open-source tool can beat any human alive, it isn’t easy to know who is playing legitimately and who is taking shortcuts.

In a strange way, that is similar to the problem facing schools and universities in the age of AI. However, chess is much farther down this road than schools, and examining what has happened to chess may provide clues on how schools can counter AI cheating.

Chess and Computers

Note: I would like to give significant credit to YouTuber Sarah Z and her excellent video, “Everyone is Cheating at Chess.” The video provides a superb overview of the chess cheating problem and serves as the inspiration for this article. 

In 1996, World Champion Garry Kasparov sat down to play the IBM supercomputer Deep Blue. It was heralded as the ultimate match of man vs. machine.

Kasparov won that six-game bout with a score of 4-2. Many breathed a sigh of relief.

However, just one year later, the two held a rematch. Kasparov noted that Deep Blue was playing significantly better, and ultimately, the computer won the series 3.5 to 2.5.

It was a significant moment. A computer defeated the world’s best human chess player. The Mechanical Turk was now real.

However, this victory was all but inevitable. Chess is a game of perfect information, meaning that all players have full knowledge of how the game is being played. There are no hidden pieces or unaccountable variables. 

In short, it’s a perfect type of problem for a computer to solve. The only thing that stopped it was the complexity of the game of chess itself, with its large number of possible moves and the need for strategic thinking. However, that was a barrier that would only hold so long, and the dam burst in 1997.

Since then, computers have only gotten more powerful. Today, an open-source app like Stockfish can easily beat any human alive. An app you can run on your phone can easily topple the best human players.

When paired with the rise of online chess, it’s easy to see why cheating and fear of being cheated are on the rise. Throw in the hyper-competitive nature of the game and content industry around chess, and you can see why the pot is boiling over.

However, schools faced their own Deep Blue moment. In November 2022, when OpenAI launched ChatGPT publicly, schools were suddenly facing the very real threat that their students could generate, rather than write, assignments.

But if chess is still struggling with its Mechanical Turk moment a quarter of a century later, what hope is there for schools? There are a few lessons to be learned.

Lesson 1: Accusations Are Easy, Supporting Them Isn’t

As Sarah Z described it in her video, cheating accusations have become a cottage industry inside the chess community. It’s ridiculously easy to accuse players of cheating, but almost impossible to prove it.

Chess.com has a massive “Cheating and Fair Play” system on its site. It claims to close over 3,500 accounts per day for fair play violations. This is achieved through various means, including comparing player moves to those of bots, identifying suspicious play patterns, and evaluating a player’s overall performance.

In more prominent tournaments, players are required to have their mic open, conduct a camera sweep of the room, and more. These protocols are similar to what students are often asked to do on remote-proctored tests.

However, these extreme measures are necessary because, while any single tool or approach can raise suspicions of cheating. It needs to be verified using other tools. 

Lesson 2: The Importance of Physical Spaces

Although most chess games are played online, major tournaments are still played over the board, also known as OTB. 

To be clear, this doesn’t eliminate cheating or accusations of cheating. The 2022 game between Hans Niemann and Magnus Carlsen proves that. However, it is much more difficult to cheat, and cheating is much less likely.

This gives OTB matches more weight, at least spiritually. Though they aren’t always practical, especially when players are geographically dispersed, the physical nature of the game makes it easier to control the environment.

Similarly, although in-class cheating does occur, it is much easier to detect and prevent. Such assignments not only need to be more common, but also weighted more heavily.

Lesson 3: The Cheat Tool as a Teaching Tool

One of the elements that Sarah Z pointed out was that Stockfish is not just the most popular way to cheat at chess, it’s also one of the most popular tools for practicing and improving at the game.

AI is somewhat similar. AI tools can be very useful for generating ideas, creating outlines, and proofreading student work. There are uses of AI that most teachers would not object to, and students would be wise to use.

But that puts the AI tool right at hand. Even as I’m writing this using Grammarly to help check for mistakes, it’s offering to “write with generative AI” and remove the need for me to write anything.

When the tool used for cheating and the tool used for learning and improvement are the same, it becomes difficult to ban the tool outright. Instead, you must establish clear boundaries around it and have a means to enforce those rules.

Lesson 4: High Pressure, Higher Stakes

Finally, one of the advantages academia has over chess is that it’s not as competitive. One student getting an A doesn’t mean another’s grade is lowered. Theoretically, all students can succeed.

But, unless you are at the very top of the chess world, the stakes are much higher in academia. Students are constantly reminded that good grades are essential to their future and, when they struggle, cheating becomes far more tempting.

That is nothing new. Contract cheating, for instance, has been a problem as long as schools have existed. AI makes it easier and thus more tempting to access tools for cheating.

Dialing down the pressure, encouraging students to make mistakes, and rewarding progress may help deter students from taking unethical shortcuts. This is especially true for younger students who are just starting on their academic journey.

Lesson 5: The Perception Problem

Finally, much like with chess, we don’t know how deep the problem of AI cheating goes in schools. Although we have multiple studies and a substantial amount of evidence indicating that it’s a pervasive issue, there’s no way to know for certain.

However, regardless of the actual severity of the problem, the perception in both cases is that cheating is rampant. That raises a simple problem: if you believe that everyone else is cheating, then it becomes easier to justify cheating yourself.

The perception can create a cycle that increases cheating by lowering the ethical threshold to do so. That increases the perception of cheating, which reduces the threshold further.

Cheating with AI is, without a doubt, a significant problem in academia, and the issue is likely even more widespread than the studies suggest. However, a perception of cheating, no matter how realistic, exacerbates the problem.

Creating a perception of fairness is a significant challenge for both the chess world and academia.

Bottom Line

Ultimately, there’s no easy solution here. If we learn one thing from chess’ struggles, it’s that they are still dealing with these issues a quarter of a century later. Computer cheating remains a significant problem with no apparent end in sight.

But there is one ray of hope. Despite these challenges, chess is still remarkably popular. According to Chess.com, they host more than 10 million games every day.  

Human beings are still playing chess and still make up the vast majority of chess games. 1997 wasn’t a catastrophe for chess, even if it opened the door to a slew of problems. 

Chess is more popular than it has been at any point in its 500+ year history. It’s not going away, despite the cheating issues.

That offers at least some hope that humans will continue to write essays in the future. The question is how to both encourage and reward authenticity while also discouraging and stopping shortcuts.

That’s a problem that even a chess grandmaster can’t easily crack.