AI Indexes
Finance AI Index
September 2026 Edition, published September 17, 2026 · The permanent record of that edition, superseded since. The current standing is on the category page →
Index › Receivables and billing › Ecommerce fraud › Small business › September 2026 Edition

Ecommerce fraud prevention for small business buyers

Asked as “ecommerce fraud prevention platform”, and as “online payment fraud detection tool”, on behalf of a small B2B company. 59 first choices recorded across the direct, paraphrase, budget and scale prompts, twelve models each. Added to the September 2026 Edition; its answers are read by the distilled judge (ai-indexes-judge-qwen3-14b-run3), not the claude-opus-5 judge of the earlier categories: how the two compare.
Standing · first-choice share
34%
Contested · Stripe Radar 20%
34NoFraud20Stripe Radar10FraudLabs Pro36others

34% of first choices, contested.

By buyer segment

The same question asked on behalf of a different buyer. Each standing is computed within its segment; they sit side by side and are never added together.

The standing

Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all twelve models, for a small B2B company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrant
01NoFraud34%0%48endorsed leader
02Stripe Radar20%0%43accepted challenger
03FraudLabs Pro10%6%16accepted challenger
04Signifyd7%34%50criticized challenger
05SEON5%5%41accepted challenger
06ClearSale5%33%18criticized challenger
07Sift2%40%42criticized challenger
Show the five products at 0%, ordered by negative rate
11Forter0%75%20criticized challenger
12Riskified0%68%28criticized challenger
10Kount0%60%20criticized challenger
09Chargeflow0%42%12criticized challenger
08Fingerprint0%24%17accepted challenger
Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its product page.

Recommended versus criticized

Every product with at least 10 labels here, on both axes. The 30% line names a quadrant, not the verdict above: that one needs more than 40%.

Criticized challengerCriticized default
Negative label rate →
01
02
03
04
05
C06
07
08
09
10
11
12
Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative50%
Key
01NoFraud34%
02Stripe Radar20%
03FraudLabs Pro10%
04Signifyd7%
05SEON5%
06ClearSale5%
07Sift2%
08Fingerprint0%
09Chargeflow0%
10Kount0%
11Forter0%
12Riskified0%

What they warned about

One of twelve models held their first choice under the paraphrase. Claude Haiku 4.5, GPT-5.4 mini, Gemini 3.5 Flash, Perplexity Sonar, Mistral Small, DeepSeek V4 Flash, Llama 4 Maverick, Qwen 3.7 Flash, Kimi K2, GLM 4.7 FlashX and MiniMax M2.5 changed. A high negative share on a product with few labels is a warning. A low share on a product with many labels is salience, not sentiment.
Riskified
68%
19 of 28 labels negative · 12 of 12 models · 4 hard negative
“Avoid Riskified unless you're enterprise-level ($50M+ revenue); their "low starting price" is misleading for SMBs.” Qwen 3.7 Flash, budget prompt
Signifyd
34%
17 of 50 labels negative · 11 of 12 models · 3 hard negative
“Users note the **high transaction fees** associated with Signifyd, which can significantly affect businesses with low margins.” Muse Glimmer 30B, negative prompt
Sift
40%
17 of 42 labels negative · 9 of 12 models · 4 hard negative
“here are platforms and approaches to avoid... **Sift** - **Issues**: Users frequently report high false positive rates” GLM 4.7 FlashX, negative prompt
Forter
75%
15 of 20 labels negative · 9 of 12 models · 3 hard negative
“**Avoid** the enterprise names (Sift, Forter, Riskified, Kount, Signifyd) for now” DeepSeek V4 Flash, budget prompt

What they cite

Citations exist only for the models that return a source list: fourteen of the twelve in this edition, and all six flagship models on the expanded tier.

Sites the answers cite

79 of 84 answers in this category came back with a source list, from 14 of 12 models: citations where the model returns them, or the search results it consulted. 1116 links across 175 sites, every framing counted. Ranked by the number of answers carrying the site or page.

vendor site · G256 answers · 151 citations · 13 models
vendor site · G242 answers · 46 citations · 13 models
vendor site · Fingerprint39 answers · 41 citations · 12 models
vendor site · Stripe36 answers · 61 citations · 12 models
vendor site · Guideflow35 answers · 36 citations · 10 models
vendor site · Fraud.net34 answers · 65 citations · 11 models
34 answers · 34 citations · 11 models
vendor site · Whop34 answers · 34 citations · 11 models
31 answers · 33 citations · 12 models
vendor site · cside25 answers · 27 citations · 11 models
vendor site · Shuftipro24 answers · 24 citations · 10 models
18 answers · 19 citations · 9 models

Pages the answers cite

The ten pages named in the most answers, by full address. A page here is one the models returned with a recommendation, not one the index endorses.

Search against answers

Each company's standing in the answers beside its site's footprint in Google search, one row a site: the products the models named on it with their shares, and the share they add up to; monthly searches on Google, and DataForSEO's estimate of AI search demand (modeled from search signals, directional, not a count of queries to any assistant), for the most-searched of the company's and its products' names (the name is in each row's hover text); estimated monthly organic visits to the site; and its best position in Google's top ten for “best ecommerce fraud prevention platform”, “ecommerce fraud prevention platform”, “ecommerce fraud prevention platforms”. US estimates from DataForSEO and Google's Ads Transparency Center. A small company's site, or a mid-sized company's site for its flagship, is marked company; a product on a large parent's site (Google, Microsoft) has no site figures. A column with no figures for this category is left out, and an empty cell means none were seen, not none exist. Two measurements side by side: neither is read as the cause of the other.
Company and productsShareOwn site citedName searches, GoogleAI search demand, est.Organic visitsPaid search a month, est.Google ads, last 30 daysBest Google position
NoFraud nofraud.com34%781,300322270
Stripe parent site
Stripe Radar 20%
20%100823,000102,741
FraudLabs Pro fraudlabspro.com10%2014008,4657
Signifyd signifyd.com7%1863,6006210,180$5,17937
SEON seon.io5%1162,9001,70924,522712
ClearSale5%4804
Shopify shopify.com
Shopify Fraud Analysis 3%, Shopify Protect 2%, Shopify Payments 0%
5%81,000,00038,0673,486,997 company$3,351110
FraudNet3%1406
Wyllo2%
Sift sift.com2%15327,1005,27316,71121

Google's top ten for the three searches hold 27 sites; 4 of them are among the sites the models cited here (clickpost.ai, fingerprint.com, learn.g2.com, stripe.com). The first result for “best ecommerce fraud prevention platform” is sift.com.

How they answered

Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.
ShowHide
ModelDirect“What is the best ecommerce fraud prevention platform for a small B2B company?”Paraphrase“Which online payment fraud detection tool would you recommend to a small business that sells to other businesses?”Comparative“What are the top ecommerce fraud prevention platforms for a small team and how do they differ?”Budget-constrained“What is the best ecommerce fraud prevention platform for a small company with a tight budget?”Scale-constrained“We are a 30 person company evaluating an ecommerce fraud prevention platform. What should we look at?”Negative“Which ecommerce fraud prevention platforms should a small business avoid or be cautious about?”
Claude Haiku 4.5FraudNet, NoFraudagainst: SignifydFraudioChanged
One alternativeSEON
against: Eftsure
NoFraud
Four alternativesClearSale, Forter, SEON, Signifyd
Wyllo
Four alternativesBolt, Fingerprint, SEON, Stripe Radar
no first choiceagainst: Feedzai, Fingerprint, Forter, Hawk, Riskified, ThreatMetrix
GPT-5.4 miniSignifyd
Two alternativesChargeflow, SEON
against: Sift
Stripe RadarChanged
Two alternativesBraintree Fraud Protection Lite, PayPal Fraud Protection
Signifyd
Five alternativesChargeflow, Fingerprint, NoFraud, Riskified, Sift
against: Kount
Stripe Radar
One alternativePayPal Fraud Protection
against: Chargeflow, Wyllo
no first choiceagainst: Fingerprint-style device intelligence tools, Riskified, Sift, Sumsub
Gemini 3.5 FlashClearSale
Four alternativesFingerprint, Resolve Pay, Signifyd, TreviPay
Stripe RadarChanged
Four alternativesClearSale, Middesk, SEON, Signifyd
NoFraud, Signifyd
Three alternativesChargeflow, ClearSale, SEON
against: FraudLabs Pro
FraudLabs Pro
Three alternativesNoFraud, Shopify Protect, Stripe Radar
ClearSale, NoFraud, Signifyd
Three alternativesSEON, Shopify Protect, Stripe Radar
against: Authorize.net, Kount, Riskified, Shopify risk filters, Sift, Signifyd
Perplexity SonarNoFraud
Two alternativesSEON, Stripe Radar
against: Forter, Riskified, Signifyd
Stripe RadarChanged
Three alternativesCybersource Fraud Management Essentials, Eftsure, FraudLabs Pro
NoFraud
Five alternativesClearSale, Kount, SEON, Sift, Stripe Radar
NoFraud
Two alternativesFingerprint, SEON
no first choiceagainst: Deep Authorization, Forter, Fraud.net, Kount, Risk Manager, Shopify Fraud Analysis, Sift, Signifyd, Visa Advanced Authorization
Grok 4.1 FastStripe Radar
Two alternativesNoFraud, SEON
against: Riskified, Sift, Signifyd
Stripe RadarHeld
Four alternativesKount, NoFraud, PayPal Fraud Protection, Sift
NoFraud, Stripe Radar
Three alternativesFingerprint, Sift, Wyllo
NoFraud, Shopify Fraud Analysis
Two alternativesStripe Radar, WooCommerce
NoFraud
Four alternativesRiskified, Sift, Signifyd, Stripe Radar
against: Chargeflow, Fingerprint, Forter, Riskified, Signifyd
Mistral SmallNoFraud, Signifyd
One alternativeSift
FraudLabs Pro, SiftChanged
Three alternativesIPQS, Mastercard Identity Check, NoFraud
Signifyd
Six alternativesChargeflow, ClearSale, Fingerprint, NoFraud, Sift, Subuno
NoFraud
Three alternativesSEON, Stripe Radar, cside
no first choiceagainst: Kount, Sift
DeepSeek V4 FlashNoFraud
Two alternativesSEON, Signifyd
against: Sift
Stripe RadarChanged
Four alternativesEftsure, Fraud.net, NoFraud, Signifyd
against: Trustmi
NoFraud
Two alternativesSEON, Signifyd
against: Forter, Kount, Riskified, Sift
FraudLabs Pro
Two alternativesNoFraud, Stripe Radar
against: Forter, Kount, Riskified, Sift, Signifyd
ClearSale, Signifyd
Two alternativesSEON, Sift
against: Forter, Kount
against: ClearSale, Forter, Kount, Riskified, Sift, Signifyd
Llama 4 MaverickNoFraud
Two alternativesFraud.net, Signifyd
FraudNetChanged
Two alternativesFingerprint, Sift
no first choiceNoFraud
Two alternativesSEON, cside
no first choicenothing named
Qwen 3.7 FlashSEON
Three alternativesFraudLabs Pro, Signifii, Stripe Radar
TrustmiChanged
Four alternativesEftsure, FraudLabs Pro, PayPal Risk Manager, Stripe Radar
Stripe Radar
Two alternativesNoFraud, Signifyd
against: Sift
Shopify Fraud Analysis, Stripe Radar
Two alternativesFraudLabs Pro, SEON
against: Riskified, Signifyd
no first choiceagainst: Forter, Ravelin, Sensfrx, Sift, Signifyd
Kimi K2NoFraud
Three alternativesChargeflow, SEON, Signifyd
Stripe RadarChanged
One alternativeSEON
NoFraud, Stripe Radar
Three alternativesFraudLabs Pro, SEON, Signifyd
FraudLabs Pro
One alternativeNoFraud
against: SEON
NoFraud, SEON
One alternativeSignifyd
against: Chargeflow, Fingerprint, Forter, Kount, Riskified, Signifyd
GLM 4.7 FlashXNoFraud
Two alternativesSEON, Signifyd
against: Sift
Stripe RadarChanged
Three alternativesEftsure, RedBlock AI, Trustmi
NoFraud
One alternativeSEON
against: Sift
FraudLabs Pro
Four alternativesChargebacks911, NoFraud, Stripe Radar, cside
against: ClearSale, Forter, Kount, Riskified, SEON, Sift, Signifyd
no first choice
Five alternativesForter, Fraud.net, Riskified, Sift, Signifyd
against: Chargeflow, Riskified, Sift
MiniMax M2.5NoFraud, SEONagainst: FraudNetNoFraudChanged
Two alternativesSEON, Signifyd
NoFraud
Four alternativesClearSale, SEON, Sift, Signifyd
NoFraud
Four alternativesSEON, Shieldtix, Stripe Radar, cside
no first choice
Four alternativesFingerprint, Sift, Signifyd, Stripe Radar
against: Chargeflow, Fingerprint, Forter, Kount, Riskified, Signifyd
Bold is the first choiceAlternatives are counted; the count opens them.What the answer argued against

The record

One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency. Full answer text is in the free responses file. Download the record
Seventy-two rows: every prompt, every model, every answer.
PromptModelVersion stringTime (UTC)SearchedSourcesLatency
Direct recommendationClaude Haiku 4.5claude-haiku-4-5-202510012026-10-04 22:21yes98 s
Direct recommendationGPT-5.4 minigpt-5.4-mini-2026-03-172026-10-04 22:30yes47 s
Direct recommendationGemini 3.5 Flashgemini-3.5-flash2026-10-04 23:12yes2534 s
Direct recommendationPerplexity Sonarsonar2026-10-04 22:45yes206 s
Direct recommendationGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-10-04 22:46yes219 s
Direct recommendationMistral Smallmistral/mistral-small via mistral2026-10-04 23:14yes138 s
Direct recommendationDeepSeek V4 Flashdeepseek/deepseek-v4-flash via deepinfra2026-10-04 23:17yes2145 s
Direct recommendationLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-10-04 22:35yes52 s
Direct recommendationQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-10-04 22:46no024 s
Direct recommendationKimi K2moonshotai/kimi-k2 via novita2026-10-04 22:18yes1723 s
Direct recommendationGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-10-04 22:42yes2127 s
Direct recommendationMiniMax M2.5minimax/minimax-m2.5 via minimax2026-10-04 22:38yes1531 s
ParaphraseClaude Haiku 4.5claude-haiku-4-5-202510012026-10-04 23:23yes97 s
ParaphraseGPT-5.4 minigpt-5.4-mini-2026-03-172026-10-04 22:14yes33 s
ParaphraseGemini 3.5 Flashgemini-3.5-flash2026-10-04 22:22yes1126 s
ParaphrasePerplexity Sonarsonar2026-10-04 22:22yes194 s
ParaphraseGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-10-04 23:10yes218 s
ParaphraseMistral Smallmistral/mistral-small via mistral2026-10-04 23:24yes85 s
ParaphraseDeepSeek V4 Flashdeepseek/deepseek-v4-flash via deepinfra2026-10-04 22:18yes2325 s
ParaphraseLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-10-04 22:59yes53 s
ParaphraseQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-10-04 22:27yes735 s
ParaphraseKimi K2moonshotai/kimi-k2 via novita2026-10-04 22:16yes2019 s
ParaphraseGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-10-04 22:59yes2440 s
ParaphraseMiniMax M2.5minimax/minimax-m2.5 via minimax2026-10-04 22:28yes1136 s
ComparativeClaude Haiku 4.5claude-haiku-4-5-202510012026-10-04 22:54yes98 s
ComparativeGPT-5.4 minigpt-5.4-mini-2026-03-172026-10-04 23:04yes76 s
ComparativeGemini 3.5 Flashgemini-3.5-flash2026-10-04 23:09yes1128 s
ComparativePerplexity Sonarsonar2026-10-04 22:40yes165 s
ComparativeGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-10-04 22:56yes1910 s
ComparativeMistral Smallmistral/mistral-small via mistral2026-10-04 22:38yes69 s
ComparativeDeepSeek V4 Flashdeepseek/deepseek-v4-flash via deepinfra2026-10-04 22:18yes2340 s
ComparativeLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-10-04 22:31yes52 s
ComparativeQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-10-04 22:48yes1841 s
ComparativeKimi K2moonshotai/kimi-k2 via novita2026-10-04 22:14yes2541 s
ComparativeGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-10-04 23:20yes2047 s
ComparativeMiniMax M2.5minimax/minimax-m2.5 via minimax2026-10-04 22:38yes1627 s
Budget constrainedClaude Haiku 4.5claude-haiku-4-5-202510012026-10-04 22:18yes1510 s
Budget constrainedGPT-5.4 minigpt-5.4-mini-2026-03-172026-10-04 22:34yes35 s
Budget constrainedGemini 3.5 Flashgemini-3.5-flash2026-10-04 23:00yes1225 s
Budget constrainedPerplexity Sonarsonar2026-10-04 22:15yes122 s
Budget constrainedGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-10-04 22:54yes2311 s
Budget constrainedMistral Smallmistral/mistral-small via mistral2026-10-04 22:21yes54 s
Budget constrainedDeepSeek V4 Flashdeepseek/deepseek-v4-flash via deepinfra2026-10-04 23:05yes2243 s
Budget constrainedLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-10-04 23:04yes53 s
Budget constrainedQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-10-04 23:10yes1021 s
Budget constrainedKimi K2moonshotai/kimi-k2 via novita2026-10-04 23:04yes2019 s
Budget constrainedGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-10-04 22:22yes2255 s
Budget constrainedMiniMax M2.5minimax/minimax-m2.5 via minimax2026-10-04 23:08yes922 s
Scale constrainedClaude Haiku 4.5claude-haiku-4-5-202510012026-10-04 23:10no08 s
Scale constrainedGPT-5.4 minigpt-5.4-mini-2026-03-172026-10-04 22:22no08 s
Scale constrainedGemini 3.5 Flashgemini-3.5-flash2026-10-04 23:21yes1325 s
Scale constrainedPerplexity Sonarsonar2026-10-04 22:33yes146 s
Scale constrainedGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-10-04 22:32yes158 s
Scale constrainedMistral Smallmistral/mistral-small via mistral2026-10-04 22:29no08 s
Scale constrainedDeepSeek V4 Flashdeepseek/deepseek-v4-flash via deepinfra2026-10-04 23:05yes1855 s
Scale constrainedLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-10-04 22:56yes53 s
Scale constrainedQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-10-04 22:30no026 s
Scale constrainedKimi K2moonshotai/kimi-k2 via novita2026-10-04 22:45yes1924 s
Scale constrainedGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-10-04 23:17yes22144 s
Scale constrainedMiniMax M2.5minimax/minimax-m2.5 via minimax2026-10-04 23:13yes1426 s
Negative framingClaude Haiku 4.5claude-haiku-4-5-202510012026-10-04 22:13yes2010 s
Negative framingGPT-5.4 minigpt-5.4-mini-2026-03-172026-10-04 22:20yes26 s
Negative framingGemini 3.5 Flashgemini-3.5-flash2026-10-04 22:46yes2827 s
Negative framingPerplexity Sonarsonar2026-10-04 23:25yes146 s
Negative framingGrok 4.1 Fastspacexai/grok-4.1-fast-non-reasoning via vertex2026-10-04 22:20yes1910 s
Negative framingMistral Smallmistral/mistral-small via mistral2026-10-04 23:25yes108 s
Negative framingDeepSeek V4 Flashdeepseek/deepseek-v4-flash via deepinfra2026-10-04 22:27yes2136 s
Negative framingLlama 4 Maverickmeta/llama-4-maverick via bedrock2026-10-04 23:20yes53 s
Negative framingQwen 3.7 Flashalibaba/qwen3.7-flash via alibaba2026-10-04 22:13yes2351 s
Negative framingKimi K2moonshotai/kimi-k2 via novita2026-10-04 22:41yes2422 s
Negative framingGLM 4.7 FlashXzai/glm-4.7-flashx via zai2026-10-04 23:09yes2233 s
Negative framingMiniMax M2.5minimax/minimax-m2.5 via minimax2026-10-04 22:53yes2234 s

Normalization in this category

Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.

ShowHide
Category-scoped readings
PayPal read as PayPal Fraud Protection
Shopify read as Shopify Fraud Analysis
Stripe read as Stripe Radar
Unresolved, counted raw
Adyen's Fraud Protection
Braintree Fraud Protection Lite
CertifID
Cybersource Fraud Management Essentials
Deep Authorization
Entrust IDV
Fingerprint-style device intelligence tools
IPQS
Mastercard Identity 360
Mastercard Identity Check (formerly Ekata)
Mastercard Identity Review 360 formerly Ekata
PayPal Risk Manager
PayPal Seller Protection
PayPal's built-in seller protection
Radar Lite
RedBlock AI
Risk Manager
Shieldtix
Shopify risk filters
Shopify's built-in fraud analysis
Signifii
ThreatMetrix
Token of Trust
Visa Advanced Authorization
WooCommerce
WooCommerce Anti-Fraud Protection
nsKnox PaymentKnox
Discontinued, still offered
No shut-down product was recommended here.
← Chargeback managementEmbedded payments →