Authority & Trust

G2's Trust and Safety Index Shows Real Scale of Review Fraud

G2's first Trust and Safety Index reveals it removed 31,014 fake reviews in a single quarter, a rare public look at review fraud happening at scale.

Black title card reading "G2's Trust and Safety Index Shows Real Scale of Review Fraud", tagged Authority & Trust, from The Pull by MagnetizeX
The short version

THE SHORT VERSION: G2 published its first Trust & Safety Index this summer, disclosing that it removed 31,014 fake reviews, 12.7% of everything submitted, between February and April 2026, alongside another 24% rejected for guideline violations. It's the clearest public data yet on how much review fraud a major B2B platform actually catches.

What happened

G2 published its Trust & Safety Index, an inaugural quarterly disclosure covering review moderation activity from February 1 through April 30, 2026. Of 244,078 reviews submitted globally, G2 published 63.3%, rejected 24% for community guideline non-compliance, and identified 31,014, or 12.7%, as fake, split between identity issues at 7.7% and incentive abuse at 5%. Every reviewer now has to authenticate through a LinkedIn account, a verified business email, or a personal email paired with a product screenshot before a review is even considered, and submissions run through automated cross-checks against millions of data points before any human moderator sees them. G2 also confirmed it has joined the Coalition for Trusted Reviews alongside Amazon, Booking.com, Expedia, Trustpilot, and Tripadvisor, an industry group focused on shared review-integrity standards.

Why review fraud data changes how buyers should read G2

From the publisher

MagnetizeX builds founder visibility systems for B2B firms.

See how →

A little over a third of everything submitted to G2 in this window never became a public review, which is a materially higher bar than most founders assume when they look at a competitor's rating. That matters two ways. For buyers, it's a reason to trust a G2 score more than an anonymous rating on a platform that doesn't publish its rejection math. For sellers, it's a warning: any shortcut involving incentivized reviews from people who can't verify a real account is now competing against a fraud-detection system built specifically to catch it, and getting caught risks a public fraud warning banner on your own listing, not just a quietly removed review.

  1. Get real customers verified, not just recruited
    G2 requires LinkedIn authentication, a verified business email, or a screenshot before a review counts. Before your next review push, confirm the customers you're asking have an easy path to one of those three, or the request will bounce at the authentication step before it ever reaches moderation.
  2. Never touch incentive abuse, even accidentally
    5% of everything rejected fell under incentive abuse specifically. A gift card tied to a guaranteed rating, rather than simply leaving a review, crosses that line. Keep any review incentive tied to submission alone, disclosed and untied to sentiment, to stay clear of the category G2 is actively enforcing against.
  3. Cite the rejection rate, not just your own score
    A 36.7% overall rejection rate is a stronger trust argument in a sales deck than your star rating alone, since it tells a skeptical buyer the number they're looking at survived real scrutiny. Reference G2's published moderation data directly when a prospect questions review authenticity.

By the numbers: G2 processed 244,078 reviews in the reporting window, publishing 154,625 (63.3%), rejecting 58,439 (24%) for guideline violations, and removing 31,014 (12.7%) as fake, including 18,735 identity-issue cases and 12,279 incentive-abuse cases.

What to do this week

If your G2 profile hasn't been claimed, claim it this week and check how many of your published reviews are more than two years old, since G2 only counts recent, first-hand experience toward its authenticity standard. Then ask three genuinely happy customers who can authenticate via LinkedIn or a work email to leave a review, rather than mass-inviting a list, since a smaller batch of verifiable reviews clears moderation more reliably than a large one that doesn't.