E-commerce Payment Fraud AI Decisions

This dataset provides granular records of e-commerce transactions flagged by an AI-powered payment fraud detection system, including transaction details, customer identifiers, risk scores, decision actions, and key model features. It enables fraud prevention teams to benchmark model performance, analyze decision rationale, and optimize fraud detection strategies.

  • last updated 5 Jan 2026
  • by GoMask
The brief that made it

Benchmarking and tuning AI fraud detection models

Sample rows

preview · 8 of 120 rows · all 22 columns
transaction_idstringdecision_actionstringtransaction_amountfloatis_flagged_fraudbooleanshipping_statestringtransaction_datetimedatetimecustomer_idstringpayment_methodstringcurrencystringshipping_street_addressstringshipping_citystringshipping_postal_codestringshipping_countrystringai_model_versionstringfraud_scorefloatdecision_reasonstringfeature_1_namestringfeature_1_valuestringfeature_2_namestringfeature_2_valuestringfeature_3_namestringfeature_3_valuestring
TXN_a93JfK1X5approve128.5falseGA2024-04-11T14:23:01CUST2031credit_cardUSD1447 Pinecrest DrAtlanta30309USmodel_v1.2.30.13low_risk_scoretransaction_frequency_30d2device_consistencytrueaddress_matchtrue
TXN-f2pK5vL9Qmanual_review2145.75trueblank2024-05-02T19:45:12CUST8874paypalEUR6 WilhelmstrasseBerlin10117DEmodel_v2.0.10.67country_mismatchdistance_km850account_countryDEprior_fraudulent_txns1
TXN_Nj7XfV8B2approve54.99falseTX2024-03-27T09:01:43CUST4092debit_cardUSD81 Elmwood LaneDallas75204USmodel_v1.2.40.11trusted_devicedevice_consistencytrueaddress_matchtrueblankblank
TXN_4dK1bZ2Lcapprove38.2falseAZ2024-04-06T17:39:20CUST2245digital_walletUSD22 Maple Ridge AvePhoenix85016USmodel_v2.0.20.18address_matchtransaction_frequency_30d4blankblankblankblank
TXN-0lK8D2zFamanual_review8720trueblank2024-05-10T12:27:33CUST1567bank_transferGBP77 Oxford StLondonW1D 2ESGBmodel_v2.1.00.81high_amountvelocity_score7200account_age_days450country_mismatchtrue
TXN_i4G7kXQ2vapprove184.95falseWA2024-04-17T10:16:48CUST3021credit_cardUSD509 Oak Tree BlvdSeattle98109USmodel_v1.3.20.09low_risk_scoretransaction_frequency_30d3device_consistencytrueblankblank
TXN-5ZRvJ7LqTreview356falseFL2024-05-13T15:52:05CUST7865paypalUSD2990 Grand AveMiami33133USmodel_v2.0.40.38recent_device_changedevice_osiOSip_country_matchtrueblankblank
TXN_1dK7cX2Lpapprove77.12falseCO2024-03-31T11:08:22CUST2378debit_cardUSD51 Cedar Park RdDenver80203USmodel_v1.2.90.16address_matchaddress_matchtrueaccount_age_days1200blankblank

What the 120 rows show

from the 120-row sample

Decline (decision action) stands out: mean transaction_amount is 30,865, against 1,879 for the rest.

  • 26%is_flagged_fraud = true
  • 126.9median transaction_amount
  • 6payment methods
  • 8feature 3 names
  • 12shipping countries
  • 13feature 2 names
Mean transaction_amount by decision_action120 rows
020k40k91.9approve60 rows451.4review33 rows30,865decline10 rows10,959manual_r…17 rows
transaction_amount120 rows, in bands of 5,000
05010099155322021025k50ktransaction_amount →

Median 126.9, from 0.15 to 45,620.

shipping_state67 rows with a value · 53 left blank
  1. TX12
  2. CA10
  3. FL4
  4. KA4
  5. AZ3
  6. WA3
  7. CO3
  8. IL3
  9. OH3
  10. MA3
22 columns by typefrom the column list below
  • string 18
  • float 2
  • datetime 1
  • boolean 1

Columns

22 columns in four groups
blueprint · 22 columns
columntypedescriptionexample
Text 18 columns
transaction_idstringUnique identifier for the e-commerce transactionuniqueTXN_a93JfK1X5
customer_idstringUnique identifier for the customer involved in the transactionCUST2031
payment_methodstringPayment method used for the transaction6 valuescredit_card
currencystringISO currency code for the transaction (e.g., USD, EUR)USD
shipping_street_addressstringStreet address for shipping destinationoptional1447 Pinecrest Dr
shipping_citystringCity for shipping destinationoptionalAtlanta
shipping_statestringState or province for shipping destinationoptionalGA
shipping_postal_codestringPostal code for shipping destinationoptional30309
shipping_countrystringCountry for shipping destination (ISO 3166-1 alpha-2 code)12 countries · optionalUS
ai_model_versionstringVersion identifier for the AI fraud detection model usedmodel_v1.2.3
decision_actionstringFinal action taken by the system for the transactionapprove · review · decline · manual_reviewapprove
decision_reasonstringShort description or code explaining the reason for the decisionoptionallow_risk_score
feature_1_namestringName of the first key feature used by the AI model for decisionoptionaldistance_km
feature_1_valuestringValue of the first key feature for this transactionoptional2
feature_2_namestringName of the second key feature used by the AI model for decisionoptionaldevice_consistency
feature_2_valuestringValue of the second key feature for this transactionoptionaltrue
feature_3_namestringName of the third key feature used by the AI model for decision8 names · optionaladdress_match
feature_3_valuestringValue of the third key feature for this transactionoptionaltrue
Numbers 2 columns
transaction_amountfloatTotal amount of the transaction in the store's currency0.01 or more128.5
fraud_scorefloatFraud risk score assigned by the AI model (0.0 to 1.0, where 1.0 is highest risk)0 to 10.13
Dates and times 1 column
transaction_datetimedatetimeDate and time when the transaction occurred2024-04-11T14:23:01
True or false 1 column
is_flagged_fraudbooleanIndicates whether the transaction was flagged as potential fraud by the AI systemfalse

Use it for

  • is flagged fra…26%31 of 120 rowsmean transaction amou…91.9appr…451.4revi…30.9kdecl…11kmanu…

    An e-commerce dashboard

    The is_flagged_fraud rate, transaction_amount by decision_action and a breakdown of shipping_state. Excel, Power BI or Tableau.

  • Why do the 10 decline rows have a mean transaction_amount of 30,865?

    A root-cause class exercise

    Hand out the rows and one question. The answer is in the data, not in the brief.

  • A software demo

    Believable transactions with transaction_datetime, customer_id and payment_method to fill a screen in front of a buyer.

Not quite right?

Make it yours.

Same 22 columns, your size and your rules. See 20 rows before you pay.

Preview 20 rows free

1,000 rows: $1.99 · 10,000: $12.99One-time. No subscription. All prices

This dataset120 rows22 columns
Yours1,000 rows22 columnsshipping_street_address: UK only

blueprint · e-commerce-payment-fraud-ai-decisions

Behind this dataset

Same schema. As many rows as you need.

These 120 rows came out of a blueprint — 22 columns with generation rules behind each one. Open it in Data Factory to retune a column, add your own, wire in foreign keys, and run it at the size you actually need.

Rules it was built with
  • Each row represents a unique payment transaction reviewed by the AI system.
  • Includes transaction features: amount, category, device type, and time of day.
  • Contains AI model score, decision (Approve/Reject/Review), and reason codes.
  • Flags for manual review and outcome (fraud confirmed or not) are recorded.
  • No personally identifiable information (PII) is included.
Rows
Open the blueprint in Data Factory

1 credit per row. New accounts start with 25 free credits.

Exports
CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
Licence
yours to use, including commercially
API slug
e-commerce-payment-fraud-ai-decisions

More in E-commerce

Browse all
Also in E-commerceSales TransactionsE-Commerce Reviews Sentiment DatasetShopping Cart Abandonment InsightsAll 1,954 datasets

What should your data show?

Preview 20 rows free
No signup. No card.