AI, Misinformation, and Why Marketers Still Have an Edge

January 3, 2026 • 3 min read • AI

I recently read an experiment from Ahrefs where they intentionally created a fake brand, seeded the internet with conflicting stories, and watched how different AI platforms responded.

The outcome was unsettling, but also oddly motivating. AI search can be manipulated more easily than most people realize, and the platforms we’ve treated as “secondary” channels are now shaping how brands are described by AI.

The goal of the experiment was simple: see how AI models handle truth when they’re given convincing lies. The result? Most AI tools confidently repeated false information if it looked detailed, credible, and well-distributed across the web, even when an official source said otherwise.

Here is what stood out to me.

Some AI platforms are far easier to manipulate than others

Not all models behaved the same. Tools like Perplexity, Grok, Copilot, and Google’s AI Overview were much more likely to accept and repeat fabricated stories once Reddit or Medium posts entered the picture.

On the other end of the spectrum, ChatGPT (specficially newer versions) was the hardest to trick. It consistently leaned on official sources when tehy existed and was far more willing to say “I don’t know” instead of inventing details.

AI has favorite “sources of truth”

Across platforms, the same domains kept showing up again and again. Wikipedia, YouTube, and Reddit were the most frequently cited overall.

But frequency isn’t the same as influence.

Reddit, Quora, and Medium punch way above their weight

When it came to steering AI responses, not just appearing in them, Reddit, Quora, and Medium had an outsized impact. A single Reddit AMA or Medium “investigation,” even when wrong, could override official brand messaging in man AI tools.

That’s a big shift from traditional search.

These platforms are no longer optional

If you’re building a brand today, Reddit, Quora, Medium, YouTube, and high-quality backlinks aren’t “nice to have” anymore. They’re part of your core marketing surface.

AI is pulling directly from these ecosystems to describe your business, whether you’re active there or not.

AIO isn’t new SEO, it’s evolved SEO

There’s a lot of noise right now around “AIO” (AI Optimization), but from everything I’ve seen, good AIO is still just good SEO with expanded inputs.

Brands that have historically invested in:

Clear site structure
Strong authority
Helpful, specific content

… are already ahead. The difference now is that off-site narratives matter more than ever, especially in places AI models trust.

Yes, this is scary, but it should also excite marketers

It’s concerning that AI could confidently cite fake companies, fake sales numbers, and fake celebrity endorsements. That part is undeniable.

But here’s the flip side: this experiment proves that search isn’t “over.” We’re not powerless in a zero-click world. Scrappy, strategic marketers still have leverage.

If your business has been hit by AI Overviews or declining clicks, this is also a blueprint. Not to spread misinformation, but to understand how narrative ownership works in AI search and to use those mechanics to grow real businesses with accurate, intentional messaging.

My takeaway

AI is going to talk about your brand whether you like it or not. The only question is whether you help shape that story, or leave it to a Reddit thread from 2019.

The future of search isn’t just rankings. It’s credibility, distribution, and making sure the most accurate version of your business is also the most viable one.

And for marketers who understand that shift early? There’s still a lot of opportunity ahead.