THIS is how to tell a genuine AI expert from a blagger who is making it up as they go along, including the “biggest red flag”.
Everyone is piling money into AI to streamline their business operations, and countless firms are suddenly “experts” in AI.
So how does a business owner know whether someone is the real deal or a blagger who learned everything they know from a bluffer’s guide?
What should you do before hiring an AI expert and what questions should you ask them to determine if they’re genuinely knowledgeable or making it all up as they go along – likely with the use of AI?
Dil Gujral, Chief AI Trainer at AI Now Academy, revealed what questions you need to ask.
He said: “When it comes to AI, experience trumps qualifications. The most important way to understand who is a genuine expert is to apply the T-N-T philosophy, namely Thinking, Not Technology. To vet their depth, see if they grasp these three critical concepts of the AI paradigm shift.
“The first critical concept to understand is that hallucinations are a feature, not a bug, and that AI is designed for creativity, not 100% factual retrieval. The second critical concept is the new AI tech we have is probabilistic and not deterministic. It’s a balance between creativity and an exact answer. The third and the most important critical concept when it comes to AI is ‘Human-in-the-Loop’.
“Humans are the pilots and, while AI proposes, humans ultimately decide.”
When it comes to AI, experience trumps qualifications
Katrina Young, Digital Transformation Strategist at KYC Digital, said you need to ask how hallucinations are controlled.
She added; “The difference isn’t just credentials, it’s how risk, data and outputs are handled. Push on risk. AI can produce incorrect outputs, so ask how results are checked, validated and corrected before use. Check data handling. Where it is stored, how it is encrypted, who has access, whether it trains external models, and what prevents leakage. Press on ethical use. How bias is identified, tested and mitigated, and how decisions remain transparent and auditable.
“Ask how hallucinations are controlled. What checks exist, what accuracy thresholds are set, and how errors are detected. Interrogate data quality. How data is sourced, cleaned, validated, versioned and audited before reaching a model. Test technical depth.
“Do they understand architecture and programming models, not just tools? Test clarity. Can they explain it in plain language? Why this approach? Ask how performance is monitored. How are outputs reviewed, drift detected, and systems retrained or reset?”
Rohit Parmar-Mistry, Founder at Burton-on-Trent-based Pattrn Data, said you need to be wary of “snake oil promises”.
He added: “We look for people who start with the business problem, not the model. We map the process end to end, then run an AI and automation audit to spot where the work is repetitive, risky or slow.
“We automate the mundane so your team has more time for judgement and customer work, not fewer people. A credible partner will also tell you where AI is not useful: weak data, unclear ownership, low stakes, or when a simple rule beats a model.
“Ask how they will provide governance and visibility: what tools are doing, logs, access controls, approvals for higher risk steps, and how issues are caught and fixed. Be wary of snake oil promises like ‘one prompt saves X hours’ or promises to replace whole roles.”
The biggest red flag is confidence without specifics
Colette Mason, Author & AI Consultant at London-based Clever Clogs AI, said a true expert should know about AI’s limitations.
She added: “Ask them what went wrong. Anyone can tell you what AI does well. That’s the sales pitch. A genuine expert will tell you about the failures, the limitations, the projects that looked brilliant on a slide deck and fell apart the moment real humans touched them.
“Ask what they built before 2022. If their entire career started when ChatGPT launched, you’re hiring someone who learned to swim last Tuesday and now offers lessons. Qualifications matter less than you’d think. There’s no single gold-standard certification. What matters is whether they can explain why something works, not just that it works.
“Can they tell you what the AI won’t do? Can they explain the trade-offs without reaching for buzzwords? The biggest red flag is confidence without specifics. Real expertise sounds like “it depends” far more often than “absolutely.” If someone guarantees AI will transform your business without first asking how you actually operate, they’re selling you a poster of a rocket, not building one.”
Mitali Deypurkaystha, AI Strategist & Author at Newcastle upon Tyne-based Impact Icon AI, said you may be speaking to an AI salesperson rather than an AI strategist.
She added: “A three-year surge in self-styled specialists has created an ‘AI adviser authenticity gap’ where story beats spin every time. The real test isn’t credentials but conversation: ask what got them into AI. Genuine experts will have a clear, experience-led answer.
“Blaggers fall back on vague tech evangelism. For me, it began with fixing my own publishing business, using AI to strip out repetitive, mundane tasks, create additional revenue streams, and make the book planning and writing process easier and faster for time-poor, purpose-led business founders. The turning point came in early 2024, when I had more enquiries from business owners seeking help with AI than from those wanting help with writing books.
“That kind of origin story shows someone who’s solved real problems under pressure, not just learned the language of the latest trend. If the answer sounds like passion for ‘innovation’ rather than proof from the coalface, you’re likely speaking to an AI salesperson, not an AI strategist.”
