Society & Ethics

Global experts back AI in education, demand exam-cheating detection

Speakers at Korea University backed AI in education while urging investment in detection tools for Korea’s CSAT, university admissions, campus exams and licensing tests.

Embrace AI, but invest in detection for high-stakes exams, global experts say amid cheating fears
Embrace AI, but invest in detection for high-stakes exams, global experts say amid cheating fearsAI-generated
By Sophie Lindqvist6 min read

Why it matters

  • At the 2026 summit, Huang said counter-cheating technology exists and funding is the main constraint.
  • Wearable-device cheating concerns extend from the CSAT to admissions, campus exams and professional licensing tests.
  • National Tsing Hua University requires students and faculty to disclose their AI use.
  • Experts proposed assessing reasoning and problem-solving rather than relying on final answers alone.

By Sophie Lindqvist

At the Times Higher Education Innovation & Impact Summit 2026 at Korea University in Seoul, global higher-education experts told The Korea Times to embrace AI as a catalyst for change while investing more in technology to detect AI-assisted cheating in Korea’s College Scholastic Ability Test (CSAT).

The proposal paired acceptance of AI with a call for stronger defenses around high-stakes exams. Experts treated those as separate jobs: teach students to use AI responsibly, and stop tools from undermining high-stakes exams where traditional testing methods are difficult to replace.

The Korea Times reported growing fears about misuse of smart glasses and other wearable devices. The concern had spread beyond the CSAT to university admissions, campus exams and professional licensing tests.

That reach makes the issue relevant to more than education departments. It touches the institutions that select students, award qualifications and maintain trust in the exams behind those decisions.

What must change for high-stakes exams?

Kuancheng Huang, a professor at National Yang Ming Chiao Tung University in Taiwan, visited Seoul to attend the summit. He put the recommendation plainly: “Just invest in technology to prevent this from happening because that's so important,” he told The Korea Times.

Huang said the technology needed to counter increasingly sophisticated cheating devices was already available. His framing makes spending the unresolved policy variable, separate from the broader need to redesign university assessment.

The immediate obstacle is how much authorities are willing to spend, not whether the tools exist, Huang said.

Why do wearables raise the stakes?

Smart glasses and other wearable devices sit at the center of the reported concern. The Korea Times said the issue had reached university admissions, campus exams and professional licensing tests, not only the CSAT.

That range spans both educational selection and professional qualification. It raises the stakes for universities and exam bodies because each setting relies on test performance to make consequential decisions.

Huang’s recommendation gave detection priority to high-stakes exams, where traditional testing methods are hardest to replace. The summit speakers did not present detection as a replacement for assessment reform.

Huang focused on protecting formal exams, while Freshwater and Kao argued that teaching, disclosure and ethics also had to change. For exams where traditional methods are difficult to replace, the proposed answer was more detection spending as cheating devices grow more sophisticated.

How should universities measure learning?

Huang suggested shifting the focus of assessments from final answers to students’ reasoning and problem-solving processes. He argued that traditional assessments are becoming less effective at measuring students’ actual abilities as AI tools grow more sophisticated.

Under that model, an assessment measures more than whether a final response is correct. It also examines the path a student used to reach that response.

Dawn Freshwater, executive director of the Worldwide Universities Network, urged universities to embrace AI as an inevitable part of education. She also said students need critical thinking and ethical judgment to use it responsibly.

She emphasized independent thinking and original work, with AI supporting rather than replacing students’ ideas. That gives assessment design a clear target: distinguish human reasoning from generated output.

The proposed measure of learning moves from the final answer alone to the student’s reasoning and problem-solving.

Why not ban AI tools?

John Kao, president of National Tsing Hua University in Taiwan, said restricting emerging technologies would do little to prevent cheating. He put the point bluntly: “Cheating has been around from the beginning of time. It's just now the tools are getting more sophisticated,” Kao said.

He continued: “It's not about taking away the tools, because the ones who cheat will always find tools to cheat with, but to emphasize academic ethics.” Kao’s position treats tool access and institutional responsibility as separate issues.

A restriction might limit some uses without removing cheaters’ access to alternatives. Ethics, in his view, has to address the behavior that restrictions alone cannot stop.

A ban alone cannot solve the problem, Kao argued; academic ethics must do the work a ban cannot.

What does disclosure change?

Kao said National Tsing Hua University requires students and faculty to disclose their AI use. He compared that practice with citing sources in academic research.

The policy makes AI use an acknowledged part of academic work. It also places a duty on faculty as well as students, because the requirement covers both groups.

Kao did not present disclosure as a prohibition. His stated aim was to make use visible while emphasizing academic ethics, the same balance Freshwater described.

Disclosure makes AI assistance visible without pretending it has no place in academic work.

What should responsible AI use look like?

Freshwater framed AI as an inevitable part of education and students’ lives. “We have to accept that students will be working and living in a world that is surrounded with AI, immersed in AI,” she said.

She said universities must teach students to think independently and produce original work. Students should use AI to support their ideas rather than replace them, she said.

She summarized the standard: “Original thought augmented by AI but not driven by AI,” she said. The phrase sets a direction for the student’s thinking and a role for the tool, rather than treating the two as interchangeable.

The distinction gives universities a practical basis for coursework and ethics instruction. It also explains why she could support both responsible AI use and stronger academic integrity.

Her standard is clear: AI should strengthen a student’s own thinking, not drive it.

What can universities act on?

The summit produced three broad directions for institutions:

  • Invest in detection for high-stakes exams, including Korea’s CSAT, following Huang’s call for greater spending on anti-cheating technology.
  • Redesign assessments around reasoning and problem-solving, following Huang’s proposal to judge the process rather than final answers alone.
  • Require transparent AI use and teach academic ethics, following Kao’s disclosure policy and Freshwater’s standard for original work supported, not driven, by AI.

These measures address different failures. Detection targets misuse in high-stakes exams, assessment reform changes what university evaluations measure, and ethics rules shape academic conduct.

For universities and authorities, the practical challenge is to act on both sides of the proposal. A detection-only strategy leaves course design unchanged, while an embrace-only strategy leaves high-stakes exams exposed to the cheating concerns Huang raised.

For Korea’s CSAT and similar exams, the immediate move is funding for detection. Across teaching and assessment, the proposed change is disclosure, ethical judgment and more attention to students’ reasoning. The next test is whether authorities and universities can deliver both without treating a ban on AI as a substitute for redesigning education.

Source: feed.koreatimes.co.kr

Share this article:

More from Sophie Lindqvist

Sophie Lindqvist

Show full bio

Staff writer covering regulation, policy and their impact on artificial intelligence and machine learning at AI In Context.

777 articles

Related articles

  1. AI Glasses Solved Four CSAT Math Questions in Ten Seconds
  2. Korea adds 8,400 staff for AI-cheating safeguards at CSAT
  3. Korea's CSAT runs under emergency anti-AI-cheating rules as minister admits detection limits
  4. AI-assisted cheater tops South Korea’s tax accountant exam
  5. South Korea Cracks Down on AI Cheating Before Thursday’s CSAT

« Previous article