AI in market research: Smarter insights or smarter fraud?

Hey there! 👋

Sharekh here! Welcome back to The Research Mag—where we break down fresh ideas, market research insights, and the innovations shaping the future of decision-making.

Before we jump into today’s discussion, let’s take a quick look at what we covered last time.

🔍 Quick recap

Last month, we explored the role of market research in Product-Led Growth (PLG) and why just having a great product isn’t enough to drive retention.

Here’s what we uncovered:

➡️ PLG isn’t just about sign-ups—users need the right onboarding, activation, and product experience to stay engaged.

➡️ Many PLG companies fail at segmentation—they assume all free trial users are potential customers, leading to high churn.

➡️ Research-driven onboarding improves retention—companies like Slack & Dropbox use deep user insights to refine activation strategies.

➡️ Competitive research matters—understanding how users evaluate alternatives helps PLG companies position themselves effectively.

PLG thrives when companies remove friction, understand user behavior, and make data-backed product decisions. If you missed it, catch up here🚀 

What’s new?

Exciting news—I’ve just launched Research Decoded with Sharekh! 🎙️ A podcast where we explore the biggest shifts in market research and decision-making.

To kick things off, I sat down with Stacy Thomas & Angie Stahl from Good Run Research & Recreation (GRRR) to tackle a critical question:

Is AI making research smarter or fueling more fraud?

🎙️ Podcast spotlight: Research Decoded with Sharekh

The first episode of Research Decoded with Sharekh is all about AI in market research—how it’s shaping insights but also introducing serious risks like fraud, fake respondents, and unreliable data.

In this episode, we unpack:

• The limits of AI-moderated research—Can automation ever match human intuition?

• Why fraud in market research is at an all-time high (and why companies are struggling to stop it).

• The future of AI + human expertise—where do we draw the line?

Here’s what came out of the conversation. 👇

🎯 Key excerpts & takeaways from the episode

1. AI moderators are efficient—but can they truly replace humans?

💬 Sharekh: “We’re seeing more AI-moderated interviews. But are they really working? Have you seen them deliver insights at the level of a human moderator?”

💬 Angie Stahl: “We’ve tested AI-moderated interviews, and here’s the reality: AI can ask questions, but it can’t think. It doesn’t follow up when something interesting comes up, it doesn’t challenge contradictions, and it certainly doesn’t pick up on emotional nuance. That’s a huge gap.”

💬 Stacy Thomas: “The best insights come when a moderator knows when to push, when to pivot, and when to dig deeper. AI isn’t there yet.”

2. The fraud problem: “People are using AI to fake expertise”

💬 Sharekh: “One of the biggest challenges we see at CleverX is that people can now fake expertise in B2B surveys. They use AI to generate responses—and it’s nearly impossible to catch unless you have the right fraud detection in place.”

💬 Stacy Thomas: “Exactly. We’ve seen ‘experts’ who can answer complex open-ended questions perfectly—but the moment you push back, they have no real knowledge. That’s because AI wrote their response, not them.”

💬 Angie Stahl: “Survey farms are getting smarter too. There are entire communities teaching people how to game research studies for money. If companies don’t start investing in better fraud detection, research is going to become completely unreliable.”

3. AI’s role in research: Assist, don’t replace

💬 Sharekh: “The way I see it—AI should be an assistant, not the decision-maker. It can help us analyze data faster, but we still need human researchers to interpret, challenge, and refine insights.”

💬 Stacy Thomas: “Exactly. The companies that win won’t be the ones that replace researchers with AI. They’ll be the ones that use AI to remove repetitive tasks, but keep humans at the center of decision-making.”

💬 Angie Stahl: “We should be using AI for data processing, fraud detection, and automation—not for deep qualitative analysis. When you need to understand emotions, motivations, or subconscious drivers, that’s where human expertise is irreplaceable.”

💬 Sharekh: “If AI-generated responses continue at this rate, the research industry faces a crisis. The integrity of our insights is at stake.”

🎧 Listen to the full episode

🎧 Spotify → Listen here

📺 YouTube → Watch the episode

🍏 Apple Podcasts → Tune in here

Fixing AI’s data problem: What comes next?

The good news? We can solve this.

✅ Stronger participant verification → Advanced fraud detection tools can flag duplicate responses, AI-generated patterns, and suspicious inconsistencies.

✅ AI-assisted fraud detection → If AI can be used to generate fake responses, it can also be used to detect them. Smart platforms are already integrating AI-powered screening techniques.

✅ Hybrid research methodologies → The future isn’t AI vs. humans—it’s AI + human expertise. Companies need both automation and human oversight to ensure data integrity.

AI is here to stay, but if companies don’t get smarter about data integrity, research will become dangerously unreliable.

So… is AI a game-changer or a data integrity nightmare?

The answer is it’s both. AI is transforming research, but without strong fraud detection and human oversight, it could cause more harm than good.

What do you think? Have you seen AI-driven research in action? Is it improving insights or making data less reliable? Hit reply and let’s talk.

📢 Stay in the loop

More episodes are coming soon—covering the biggest challenges and innovations in market research. 🔗 Subscribe Now

Let’s keep the conversation going. I’d love to hear what you think! Got an idea for a future topic? Reply to this email and let’s talk.

You can also reach out to me directly here.

Sharekh,
The Research Mag
Founder @CleverX
Connect with me on X and LinkedIn