June 20, 2025
AI
85% of teachers and 86% of students used AI in the 2024–25 school year, according to “Schools’ Embrace of AI Connected to Increased Risks,” a report from the Center for Democracy and Technology.
A 2024 study in Frontiers in Psychology also found that nearly 1 in 5 students reported using AI tools during graded work without instructor permission. The actual number is likely much higher.
As AI becomes more integrated into everyday learning, its use during assessments introduces real challenges. Students now use tools like ChatGPT and Gemini not just for studying, but also during exams, sometimes in ways that are hard to detect. Traditional monitoring methods aren’t keeping pace.
Trying to manage AI-assisted misconduct with manual review alone is like bringing a bicycle to the Indy 500. It’s simply not fast or sophisticated enough to keep up with the tools students are using.
That’s where AI proctoring comes in: a way to help instructors identify irregular activity during online assessments without relying on guesswork or over-surveillance.
What AI Proctoring Does
AI proctoring uses machine learning to support exam integrity. It monitors assessments for patterns or behaviors that may indicate unauthorized assistance, such as multiple faces on screen, a test-taker leaving the frame, or signs of another device being used.
The system flags these instances for instructors to review. It does not make decisions or issue penalties; it provides information so educators can determine what’s appropriate based on context.
How AI Is Being Used During Exams
Students now have access to a range of AI and connected tools during assessments, including:
- Generative AI platforms that provide real-time answers
- Smart devices like glasses or earbuds
- Live-assistance apps, such as Cluely
- Browser extensions that transmit exam content to outside platforms
These tools make it increasingly difficult to draw a clear line between legitimate use and academic misconduct, particularly in remote or unsupervised environments.
Why Manual Oversight Falls Short
Human oversight remains important but has clear limitations:
- A single exam can produce hours of footage per student
- The signs of misconduct are often subtle or invisible without assistance
- Reviewers may misinterpret natural behavior, especially under pressure or time constraints
AI proctoring helps by filtering large amounts of data and highlighting patterns that warrant closer inspection, enabling a more focused and efficient review process.
Key Features of AI Proctoring
AI proctoring typically supports academic integrity in two ways:
Behavioral Monitoring
Identifies visual or audio cues that suggest potential issues, such as a missing face, the sudden appearance of a phone, or eye movements consistent with external prompting.Detection of Unauthorized Tools
Recognizes when banned applications are launched, AI platforms are accessed, or physical devices appear within the camera view.
These insights are surfaced to instructors for review, not for automatic enforcement.
Privacy, Fairness, and Transparency
AI proctoring should be implemented with clear ethical boundaries:
- Transparency: Students and instructors are informed about what is monitored
- Human judgment: Final decisions are always made by people, not algorithms
- Data privacy: All data is encrypted, limited in scope, and stored securely by the institution
The purpose of AI proctoring is not surveillance, it’s to ensure consistent testing conditions and protect academic standards. That includes respecting student privacy and giving institutions full control over how data is used.
Bottom Line
AI is now a regular part of education, for both students and instructors. As its use grows, so does the potential for misuse during assessments.
Manual oversight alone is no longer enough. AI proctoring offers a practical way to support fair, secure, and scalable assessment processes, while keeping people, not algorithms, in control.
Because when the pace of change accelerates, you can’t bring a bicycle to the Indy 500.
