Is cheating becoming OK? What a viral Reddit post reveals about AI and interviews
A recent Reddit thread asks a blunt question: is cheating now acceptable when AI tools are everywhere? The poster, a 32-year-old who casually uses ChatGPT, says their 21-year-old nephew “cheated on all his interviews” to land an offer at a company the poster previously worked with. The OP also referred the candidate to HR.
“My nephew 21M recently got an offer… and my sister told me that he cheated on all his interviews.”
It’s a timely dilemma. AI is now part of mainstream work, from drafting emails to writing code. But where is the line between smart preparation and dishonest advantage in hiring? And what should UK employers and candidates do differently?
Here’s a practical look at what the post means, why it matters, and how to respond.
Where is the line: AI-assisted vs dishonest interview behaviour
Using AI to prepare isn’t new – we’ve all used notes, templates, or coaching. What’s changed is the ease of real-time assistance. With a browser side-panel, AI can suggest answers live, summarise technical questions, and even generate code or strategy on the fly.
Reasonable AI use (generally acceptable)
- Research: using a large language model (LLM – a text-generating AI) to study a company, role, or core concepts.
- Practice: running mock interviews, refining STAR answers, or tightening your CV.
- Drafting: improving the clarity of your portfolio, cover letters, or take-home write-ups.
Crossing the line (likely cheating)
- Real-time, undisclosed AI answering during interviews or assessments intended to test your own knowledge.
- Submitting AI-generated code, analysis, or writing as your own without permission where originality is being assessed.
- Claiming experience or qualifications you don’t have, AI-assisted or not.
Ethically, the test is simple: if the assessor reasonably expects personal competence, undisclosed AI help becomes deception.
Why this matters in the UK: ethics, law, and practical risk
UK law doesn’t have a special category for AI-enabled cheating, but general principles apply:
- Fraud Act 2006: knowingly making false representations to gain employment can constitute fraud by false representation. See the legislation: Fraud Act 2006, Section 2.
- Employment consequences: misrepresentation discovered post-hire can justify rescinding an offer or dismissal for gross misconduct.
- Data protection: employers considering monitoring (e.g., proctoring, keystroke logging) must comply with UK GDPR and be proportionate. See the ICO’s guidance on employment practices and AI: Employment practices and AI and data protection.
There’s also the trust factor. Even if AI-enabled cheating isn’t criminal, it shakes confidence in hiring and can damage a referrer’s reputation. A personal referral is fine; undisclosed assistance is not.
What the Reddit post tells us about workplace reality
The OP isn’t an “AI person”, but uses ChatGPT like millions do. That’s the point: the barrier to cheating is low, and tools are invisible. We can’t assume that traditional interview formats still measure what we think they do, especially over video.
Employers need to design AI-aware assessments. Candidates need clarity: when is AI helpful, and when is it dishonest?
Practical guidance for candidates: use AI without crossing the line
- Prepare with AI, perform yourself: study, simulate, and polish with an LLM before the interview; don’t outsource your answers during it.
- Be transparent: if you used AI to draft a portfolio or take-home, say so and explain your process, judgement, and what you learned.
- Clarify “allowed tools”: ask recruiters what’s permitted for take-home tasks. Some allow general tooling; others expect solo work.
- Build durable competence: AI can coach you, but you still need core skills. If a live pairing test is required, you’ll need to demonstrate them.
- Own your growth areas: if AI helped you grasp a concept, credit it and show how you’d handle similar work independently.
If you want to use AI to speed up legitimate admin tasks, that’s a different story – and a good one. For example, here’s how to connect ChatGPT to Google Sheets with a Custom GPT to automate reporting and avoid copy-paste drudgery.
Actionable steps for UK employers: fair, AI-aware hiring
Design assessments that reward real capability
- Use structured interviews: consistent questions and scoring rubrics reduce noise and make gaming harder.
- Prefer work-sample tests: short, supervised exercises (pair programming, case walkthroughs) closely mimic the job and limit covert assistance.
- Assess judgement, not just recall: ask “how you’d approach X” and “trade-offs you’d consider”. AI-generated answers often falter on specifics and constraints.
Be explicit about permitted AI use
- Publish a “tooling policy” for interviews and take-homes. State what’s allowed (e.g., IDEs, documentation) and what isn’t (live LLM answers).
- Include a candidate declaration for take-home tasks. Keep it light-touch and non-accusatory.
- Offer an AI-disclosure option. Many candidates will be honest if the norm is clear and fair.
Use proportionate safeguards and respect privacy
- Avoid invasive monitoring by default. If you use proctoring, conduct a Data Protection Impact Assessment (DPIA), update your privacy notice, and limit data collected. See the ICO’s employment guidance linked above.
- Be cautious with “AI detectors”. They are unreliable and risk bias. Treat them as weak signals, not evidence.
- Confirm skills post-hire: probationary projects, code reviews, and mentoring surface gaps early without assuming bad faith.
Red flags to consider – and benign explanations
- Overly polished, generic phrasing repeated across answers. Could be AI – or just a well-prepped candidate.
- Long pauses, then sudden fluent responses. Could be reading from a script – or taking notes thoughtfully.
- Portfolio that doesn’t match live competence. Probe with specifics: decisions, constraints, metrics, and lessons learned.
Always give room for alternative explanations. Inclusive hiring means testing fairly and avoiding snap judgements, especially with candidates for whom English is a second language.
Where I land on the Reddit dilemma
Referring a family member for an interview is common and not inherently unfair. Cheating through interviews is different. It’s a breach of trust with the team, the referrer, and the employer, and can backfire quickly once the real work starts.
AI isn’t the villain here; unclear norms are. Employers should update processes for an AI-native world. Candidates should use AI to learn and prepare, then show their own judgement when it counts.
Further reading and sources
- Original Reddit thread: Is cheating becoming OK?
- ICO – Employment practices and data protection: Guidance
- ICO – AI and data protection: Guidance
- Fraud Act 2006 – False representation: Legislation
- CIPD – Recruitment: an overview of good practice: Factsheet
The question isn’t “is cheating becoming OK?” It’s “how do we make AI use transparent and fair so we can focus on real capability?” That’s a norm we can all get behind.
