AI Cheating in Recruitment: What HR Needs to Know

With the rise of generative AI tools like ChatGPT, concerns about cheating in recruitment—particularly in online assessments—are louder than ever.Can candidates simply ask an AI to do the test for them? And if so, does that undermine the value of psychometric testing?

At Test Partnership, we take these concerns seriously. But the real picture is more nuanced than the headlines suggest.

With the right approach, personality tests can be both fair and inclusive—even for neurodivergent applicants.

The uncomfortable truth is this: generative AI can now outperform most human candidates on ability tests. It can solve numerical reasoning questions, understand complex verbal analogies, and even complete logical puzzles at a high level. In unsupervised settings, AI can easily be used to simulate top-tier performance.

The same is true of application forms and CVs. What was once a useful way to gauge motivation or communication style is now often a product of ChatGPT. In fact, organisations relying heavily on these inputs are increasingly rewarding prompt engineering rather than actual candidate quality.

Put simply, AI hasn’t just crept into recruitment—it’s transformed the rules entirely.

The priority is not preventing AI use—it’s designing a hiring process that acknowledges and adapts to this new reality. Trying to “catch” AI is a losing game. Instead, employers need to re-evaluate the predictive value and integrity of each selection tool in light of AI’s capabilities.

At Test Partnership, we recommend:

• Shifting from free-text to structured formats – Replace written answers and CVs with structured, quantifiable assessments that measure job-relevant traits more directly.
• Re-prioritising supervised assessments – If you're hiring for roles where integrity and genuine ability are crucial, consider assessments under supervision or proctored conditions—virtual or in-person.
• Focusing on traits AI can't fake – Traits like learning potential, behavioural tendencies, and job-relevant decision-making are less susceptible to AI assistance and can be reliably measured through well-constructed psychometric tools.
• Transparent candidate communication – Be upfront with candidates about the role of AI, fairness, and the need for authenticity in the process. Good candidates will respect that.
• Rethinking what “cheating” means – If everyone uses AI to draft a CV or answer application questions, is it cheating—or just adaptation? The solution is not tighter control, but better-designed assessments.

Only if the system isn’t built to handle it.

At Test Partnership, we constantly review how AI tools interact with assessments—and we evolve accordingly. More importantly, we focus on what matters: ensuring every candidate is measured fairly, accurately, and under conditions that reflect real-world ability—not just access to tech.

Yes, AI is powerful. But with the right design, policies, and technology, the threat of AI-assisted cheating in recruitment can be managed—and minimised.

Want to learn how Test Partnership protects your hiring process from AI abuse? Get in touch with our team today.