I’ve spent fifteen years in eCommerce, watching businesses navigate the messy reality of customer feedback. I’ve seen the good, the strategic, and the outright fraudulent.
Here’s what I know: the shortcuts that worked in 2020 are becoming liabilities in 2025.
AI doesn’t read reviews the way humans do. It doesn’t scan for keywords or get fooled by enthusiastic language. It synthesises patterns across millions of data points, detects inconsistencies that no human team could spot, and surfaces what’s real with increasing precision.
The businesses that gamed the old system are about to have their shortcuts exposed.
The Detection Leap You Didn’t See Coming
Google blocked over 240 million policy-violating reviews in 2024.
That number alone should make you pause. But here’s the part that matters more: the accuracy of identifying fake reviews improved by 45% compared to 2022.
This isn’t incremental progress. This is a fundamental shift in what platforms can detect.
Google trained a new model with the help of Gemini that identifies potentially suspicious profile edits. The system regularly analyses reviews and detects new patterns of abuse, even months after they’ve been posted.
AI doesn’t just catch fraud in real-time. It retrospectively hunts down manipulation that previously went undetected.
That review campaign you ran in 2022? The one where you offered discounts for five-star reviews? AI is looking at it now, finding patterns in timing, language, and reviewer behaviour that human moderators missed.
The Permanent Consequences Are Already Here
I’ve watched businesses lose something they can’t get back: the ability to collect Google reviews.
Not temporarily suspended. Permanently gone.
No appeals. No second chances. Finished.
Google puts explicit statements on your listing telling every potential customer you were caught buying fake reviews. Not vague warning labels. Direct public accusations visible to everyone searching for your business.
The most alarming part? Google’s looking back at historical patterns going years. Bought fake reviews back in 2022 or 2023? You could suddenly face consequences in 2026 for something you did years ago.
The statute of limitations on review manipulation is effectively gone.
How AI Actually Sees Through the Games
Amazon uses deep graph neural networks to analyse complex relationships and behaviour patterns. The platform blocked over 250 million suspected fake reviews in 2023 alone.
Machine learning models analyse whether the seller has invested in ads, customer-submitted reports of abuse, risky behavioural patterns, and review history. Large language models work alongside natural language processing techniques to analyse anomalies that might indicate a review is fake or incentivised.
The system goes beyond surface-level indicators to identify deeper relationships between bad actors.
Here’s what AI detects that humans miss:
Pattern clusters. When multiple reviews come from accounts created within days of each other, using similar language structures, posted within narrow time windows.
Location inconsistencies. AI uses location data to verify whether reviewers were actually present at the businesses they critique. This distinguishes between authentic and fabricated experiences.
Network relationships. The connections between reviewer accounts, identifying coordinated campaigns that look organic on the surface but reveal manipulation when analysed as a network.
Language anomalies. AI-generated fake reviews demonstrate higher comprehensibility and lower levels of specificity, exaggeration, and negligence compared to authentic reviews. They exhibit significantly higher mechanicalness and lower empathy.
Advanced algorithms excel at spotting outliers and deviations from typical review patterns, flagging potentially fraudulent activities that blend in when viewed individually.
The Human Detection Failure
Research shows that humans have 57% accuracy in a fake review detection task.
That’s barely better than a coin flip.
Fake reviews have become more sophisticated than ever before. Bots, fake accounts, and networks of people paid to leave fake reviews make it easier to create convincing reviews faster than a human team can realistically keep up with.
This is why businesses that relied on manual detection or assumed platforms wouldn’t catch manipulation are now exposed. The gap between human capability and AI detection has widened to the point where shortcuts are systematically revealed.
The Legal Reckoning Adds Weight
Rule violations can result in civil penalties of up to $53,088 per violation.
In December 2025, Federal Trade Commission staff sent letters to 10 companies, warning them of potential violations of the agency’s Consumer Review Rule. The rule prohibits certain deceptive or unfair conduct related to the use of product reviews in advertising and marketing.
These letters confirm that companies using fake reviews or providing incentives for five-star reviews misrepresent consumer experiences and opinions. They may be subject to FTC enforcement actions and civil penalties.
The combination of platform detection and regulatory enforcement creates a pincer movement that businesses can’t dodge.
What This Means for Australian Businesses
I work with eCommerce businesses, trades, home services, retailers, hospitality, and professional services across Australia. The conversation has shifted in the past year.
Businesses are asking different questions now. Not “How do we get more reviews?” but “How do we make sure our review process is bulletproof?”
The fear is real. The consequences are visible. The old playbook doesn’t work.
Here’s what I tell them:
Authenticity isn’t a nice-to-have anymore. It’s survival.
Real feedback from real people builds lasting reputations. The businesses that understand this are building systems that capture genuine customer experiences at the right moments, without manipulation or shortcuts.
Transparency protects you.
Clear, honest communication with customers about why reviews matter and how you use feedback creates trust. AI can detect when businesses are being straightforward versus when they’re gaming the system.
Systems matter more than tactics.
Automation and structure simplify the review process without crossing ethical lines. You can make it effortless for happy customers to share their experiences without incentivising or manipulating outcomes.
The Window Is Closing
AI detection improves every quarter. What platforms miss today, they catch tomorrow. What they catch tomorrow, they apply retrospectively to historical data.
Businesses that built their reputation on fake reviews are watching that foundation crumble. The public accusations on Google listings, the permanent loss of review collection, the FTC penalties—these aren’t theoretical risks anymore.
The reckoning is happening now.
If you’ve been relying on shortcuts, you have two choices: clean up your review process before AI finds the patterns, or wait for the consequences to find you.
I’ve seen both paths. One involves proactive system building, authentic customer engagement, and sustainable reputation management. The other involves damage control, lost trust, and permanent platform restrictions.
The businesses that survive this shift are the ones that realise reviews aren’t a marketing tactic. They’re a reflection of how you actually treat customers.
AI just makes that reflection impossible to fake.
Building the Right System
At Gather More Reviews, we help Australian businesses turn happy customers into powerful marketing by making it effortless to capture, manage, and showcase reviews that drive trust, traffic, and growth.
We don’t help you game the system. We help you build one that works with AI detection, not against it.
Our approach focuses on timing, authenticity, and systematic feedback collection that platforms recognise as legitimate. We blend technology with empathy, strategy with simplicity, and data with common sense.
Whether you’re a solo tradie or a national retail chain, the same principle applies: make it easy for genuine customers to share genuine experiences.
That’s the only approach that survives what’s coming.
If you want to build a review system that works in 2026 and beyond, let’s chat:
The businesses that get this right now will have a competitive advantage that compounds. The ones that wait will spend years recovering from shortcuts that seemed smart at the time.
AI doesn’t forgive. It just exposes.
Choose accordingly.
