TAIPEI, Taiwan — Artificial intelligence is delivering a serious blow to literature and the people tasked with judging it, Taiwanese writer Chu Yu-hsun has warned, saying that nearly 40 percent of novel submissions he reviewed this year appeared to be AI-generated.

Chu said the growing flood of AI-written entries has transformed literary judging into what he described as one of the most difficult experiences of his career.

Writing on Facebook on Saturday, Chu said the literary community is now trying to defend human creativity against what he described as an almost infinitely replicable industrial production line powered by artificial intelligence.

And the problem, he warned, could eventually threaten the entire literary award system.

Nearly 40% of Submissions Appeared to Be AI-Generated

Chu said he began noticing a disturbing pattern while reviewing batches of submissions.

At first, he did not always immediately realize that the entries were generated by AI. However, after reading multiple works, he began recognizing repeated writing styles and similar patterns that suggested the use of increasingly sophisticated AI models.

He said there was also a noticeable difference between AI-generated works submitted during the first and second halves of the year, suggesting that the technology was evolving rapidly.

The change has fundamentally altered the job of literary judges.

Previously, judges could focus primarily on identifying the strongest and most original works.

Now, they must also try to determine a more basic question:

Was this written by a human at all?

‘Like Facing Three Balls at Once’

Chu used a baseball analogy to describe the challenge facing literary judges.

He said reviewing submissions had become like facing a pitcher throwing three balls at the same time — with two of them being illusions.

Before deciding whether to swing, the batter must first determine which ball is real.

For literary judges, the problem is similar.

They must evaluate the quality of a work while simultaneously trying to identify whether artificial intelligence played a role in creating it.

Chu said the additional burden can dramatically increase the amount of work required.

Reviewing 200 submissions that include 20 suspected AI-generated works, he said, could require as much effort as reviewing 400 or even 500 conventional submissions.

The current system, he warned, is placing enormous pressure on judges while offering no reliable way to guarantee that AI-generated content will be detected.

Even a 95% Accuracy Rate Could Mean Dozens of Mistakes

The problem is not simply identifying suspicious writing.

AI detection can also produce serious mistakes.

Chu warned that judges could incorrectly identify human-written works as AI-generated while failing to detect works that were actually created by artificial intelligence.

Even with an accuracy rate of 95 percent, he said, reviewing 1,000 submissions could still result in 50 incorrect judgments.

That margin of error could have serious consequences for writers competing for recognition, prize money and opportunities to build their careers.

The more AI-generated submissions enter literary competitions, Chu argued, the harder it becomes to ensure that genuine human writers are judged fairly.

‘The Literary Award System Could Collapse’

Chu issued one of his strongest warnings when discussing the future of literary awards.

He said AI-generated submissions could eventually destroy the existing award system if organizers fail to introduce stronger protections.

The workload for judges has deteriorated to such an extent, he said, that compensation would need to increase substantially if literary awards want to maintain the quality of their judging.

Otherwise, he warned, competitions could become increasingly vulnerable to mistakes and disputes.

For Chu, the issue is no longer a distant concern about the future of technology.

It is already affecting literary competitions now.

Writer Says He Will Refuse to Judge Some Competitions

Chu said that beginning next year, he will refuse to serve as a judge for literary awards that do not explicitly prohibit AI-generated submissions.

Instead, he said he would prioritize competitions that introduce stronger measures to protect the integrity of human-created work.

Among the measures he suggested are:

• Requiring contestants to sign declarations confirming the originality of their work
• Asking writers to preserve records of their creative process
• Introducing clearer rules explicitly prohibiting AI-generated entries
• Creating stronger systems for investigating suspicious submissions

The goal, he said, is not simply to punish people who use AI.

It is to protect the credibility of literary awards and ensure that genuine writers still have a meaningful opportunity to be discovered.

A Warning to Writers Using AI

Chu also directed a sharp warning at contestants who use artificial intelligence to create entries and submit them as their own work.

He argued that people exploiting weaknesses in the current system may eventually discover that they have destroyed the very platform they depend on for recognition and prize money.

Once literary award systems lose credibility or collapse under the pressure of AI-generated submissions, he warned, writers who depend on those competitions could find themselves with fewer opportunities to have their work recognized.

The debate raises a much larger question for the literary world:

How can competitions embrace technological change without allowing artificial intelligence to overwhelm human creativity?

AI Is Changing the Future of Literature

Artificial intelligence is rapidly transforming creative industries around the world.

AI tools can now generate essays, stories, poetry, scripts and other forms of written content within seconds.

Supporters argue that the technology can help writers brainstorm ideas, edit drafts and experiment with new forms of creativity.

Critics, however, fear that unrestricted AI-generated content could make it increasingly difficult to distinguish between human and machine-created work.

For literary competitions, the challenge is particularly serious because originality is often at the heart of the judging process.

If judges cannot reliably determine who — or what — created a submission, the entire foundation of a competition can come into question.

The Fight to Protect Human Creativity

Chu’s warning reflects a growing concern within creative communities:

Artificial intelligence may be able to generate an endless amount of content, but literary institutions still need a way to identify, reward and protect human creativity.

For now, literary judges are being asked to do more than evaluate writing.

They are becoming investigators as well.

And unless competitions quickly adapt, Chu warned, the flood of AI-generated content could overwhelm a system that was built on the assumption that every submission came from a human author.

The question facing the literary world is no longer whether AI will change writing. It already has.

The real question now is whether literary awards can adapt quickly enough to protect the writers they were created to celebrate.

WWC ONE MEDIA J.M.D