Insurance
23 columns
200 rows
8 formats

Insurance Claim Fraud Scenarios

This dataset contains synthetic insurance claim records enriched with detailed event features and labeled fraud scenarios, designed for advanced machine learning-based fraud detection. Each record includes customer, policy, incident, and claim details, along with structured fraud indicators and event-level metadata, making it ideal for developing and benchmarking fraud detection algorithms.

last updated 29 Jul 2025
by GoMask

The brief that made it

Training and benchmarking ML models for insurance fraud detection

Sample rows

preview · 8 of 200 rows · all 23 columns

claim_idstring	claim_statusstring	claim_amountfloat	fraud_labelboolean	incident_typestring	policy_idstring	customer_idstring	claim_datedate	incident_datedate	incident_location_streetstring	incident_location_citystring	incident_location_statestring	incident_location_postal_codestring	incident_location_countrystring	fraud_scenariostring	reported_delay_daysinteger	number_of_previous_claimsinteger	policy_start_datedate	policy_end_datedate	customer_ageinteger	customer_genderstring	customer_occupationstring	event_featuresstring
CLM00001	approved	1750.45	false	collision	PLC10001	CST90001	2024-03-05	2024-03-03	1200 Market St	San Francisco	CA	94103	USA	blank	2	0	2022-10-01	2025-09-30	29	female	designer	{"impact_speed":26,"police_report":true,"damage_estimate":1600}
CLM00002	paid	670	false	theft	PLC10002	CST90002	2024-01-18	2024-01-15	45 Pine Ave	Chicago	IL	60611	USA	blank	3	2	2023-02-14	2025-02-13	34	male	engineer	{"stolen_item":"laptop","police_report":true,"witness_count":0}
CLM00003	approved	38000	false	fire	PLC10003	CST90003	2022-06-21	2022-06-20	789 Maple Blvd	Tulsa	OK	74105	USA	blank	1	5	2019-01-01	2024-12-31	62	female	retired	{"fire_department_report":true,"damage_level":"major","injuries":0}
CLM00004	submitted	550	false	collision	PLC10004	CST90004	2023-07-10	2023-07-08	22 Grand Ave	New York	NY	10013	USA	blank	2	0	2021-12-01	2024-11-30	24	female	student	{"impact_speed":10,"police_report":false,"damage_estimate":500}
CLM00005	approved	9000	false	theft	PLC10005	CST90005	2024-04-18	2024-04-17	88 Ocean Dr	Miami	FL	33139	USA	blank	1	1	2023-05-01	2026-04-30	28	male	investment_banker	{"stolen_item":"bicycle","police_report":true,"witness_count":2}
CLM00006	approved	22000	false	natural_disaster	PLC10006	CST90006	2023-11-22	2023-11-15	15 Hillcrest Rd	San Diego	CA	92103	USA	blank	7	6	2022-01-01	2024-12-31	70	male	pensioner	{"event_type":"flood","damage_level":"moderate","evacuation":true}
CLM00007	under_review	1200	false	collision	PLC10007	CST90007	2023-05-12	2023-05-11	2458 Roosevelt Rd	Phoenix	AZ	85009	USA	blank	1	1	2022-06-10	2024-06-09	26	female	bartender	{"impact_speed":34,"police_report":true,"damage_estimate":1000}
CLM00008	paid	18000	false	fire	PLC10008	CST90008	2024-01-05	2023-12-30	3202 Willow Ln	San Jose	CA	95128	USA	blank	6	5	2021-07-01	2024-06-30	59	male	business_owner	{"fire_department_report":true,"damage_level":"moderate","injuries":1}
CLM00009	under_review	7100	false	collision	PLC10009	CST90009	2023-09-14	2023-09-14	1800 Main St	Dallas	TX	75201	USA	blank	0	3	2022-09-10	2025-09-09	41	female	teacher	{"impact_speed":22,"police_report":true,"damage_estimate":6500}
CLM00010	approved	3200	false	vandalism	PLC10010	CST90010	2023-03-02	2023-02-28	501 Spruce St	Denver	CO	80209	USA	blank	2	0	2021-03-01	2024-02-28	36	male	graphic_designer	{"vandalism_type":"graffiti","police_report":true,"repair_estimate":3000}
CLM00011	submitted	20000	false	collision	PLC10011	CST90011	2023-02-10	2023-02-09	3300 Lakeview Dr	Los Angeles	CA	90026	USA	blank	1	1	2022-02-01	2025-01-31	31	female	real_estate_agent	{"impact_speed":45,"police_report":true,"damage_estimate":18500}
CLM00012	submitted	670	false	collision	PLC10012	CST90012	2022-08-15	2022-08-14	810 Valley Rd	Atlanta	GA	30303	USA	blank	1	0	2020-08-01	2023-07-31	22	female	student	{"impact_speed":18,"police_report":false,"damage_estimate":600}
CLM00013	approved	2250	false	theft	PLC10013	CST90013	2023-01-30	2023-01-29	910 4th St	Seattle	WA	98101	USA	blank	1	2	2022-01-01	2025-12-31	27	male	software_engineer	{"stolen_item":"phone","police_report":true,"witness_count":1}
CLM00014	paid	13000	false	natural_disaster	PLC10014	CST90014	2023-10-02	2023-09-30	600 Oak Ave	Houston	TX	77002	USA	blank	2	8	2018-10-01	2026-09-30	67	female	retired	{"event_type":"tornado","damage_level":"major","evacuation":false}
CLM00015	submitted	250	false	theft	PLC10015	CST90015	2024-02-22	2024-02-21	2200 2nd Ave	Detroit	MI	48201	USA	blank	1	0	2023-01-01	2025-12-31	25	male	student	{"stolen_item":"wallet","police_report":false,"witness_count":0}
CLM00016	under_review	19999.99	false	fire	PLC10016	CST90016	2023-04-25	2023-04-21	1870 Cedar Rd	Las Vegas	NV	89109	USA	blank	4	4	2021-04-15	2024-04-14	64	male	business_owner	{"fire_department_report":true,"damage_level":"minor","injuries":0}
CLM00017	approved	995	false	collision	PLC10017	CST90017	2024-05-01	2024-04-30	4200 Ash St	Portland	OR	97205	USA	blank	1	0	2022-05-01	2025-04-30	30	female	graphic_designer	{"impact_speed":16,"police_report":false,"damage_estimate":950}
CLM00018	approved	20500	false	natural_disaster	PLC10018	CST90018	2023-12-03	2023-12-01	1099 Elm St	Minneapolis	MN	55402	USA	blank	2	9	2020-12-01	2025-11-30	74	female	retired	{"event_type":"hailstorm","damage_level":"minor","evacuation":false}
CLM00019	approved	1500	false	theft	PLC10019	CST90019	2022-10-13	2022-10-11	2105 Birch Dr	Austin	TX	78701	USA	blank	2	2	2021-10-01	2024-09-30	36	male	teacher	{"stolen_item":"camera","police_report":true,"witness_count":0}
CLM00020	rejected	0.01	true	other	PLC10020	CST90020	2022-11-17	2022-11-17	blank	blank	blank	blank	blank	inflated_claim	0	0	blank	blank	0	blank	blank	{"description":"minimal info","discrepancy":true}

What the 200 rows show

from the 200-row sample

Rejected (claim status) stands out: 36 of its 37 rows have fraud_label = true, against 22 of 162 for the rest.

29%fraud_label = true
1,800median claim_amount
3incident location countries
3customer genders
6fraud scenarios
25customer occupations

Fraud label rate by claim_statusfraud_label = true

claim_amount199 rows, in bands of 100k

Median 1,800, from 0.0 to 1,000,000.

incident_type199 rows with a value · 1 left blank

collision63
theft48
natural_disaster27
fire26
vandalism18
other17

23 columns by typefrom the column list below

string 14
integer 3
float 1
date 4
boolean 1

Columns

23 columns in four groups

blueprint · 23 columns
column	type	description	example
Text 14 columns
`claim_id`	string	Unique identifier for each insurance claim recordunique	CLM00001
`policy_id`	string	Unique identifier for the insurance policy associated with the claim	PLC10001
`customer_id`	string	Unique identifier for the customer who filed the claim	CST90001
`incident_type`	string	Type of incident (e.g., collision, theft, fire, natural disaster)6 values	collision
`incident_location_street`	string	Street address where the incident occurredoptional	1200 Market St
`incident_location_city`	string	City where the incident occurredoptional	San Francisco
`incident_location_state`	string	State or province where the incident occurredoptional	CA
`incident_location_postal_code`	string	Postal code of the incident locationoptional	94103
`incident_location_country`	string	Country where the incident occurred3 countries · optional	USA
`claim_status`	string	Current status of the claim (e.g., submitted, approved, rejected, under_review, paid)submitted · approved · rejected · under_review · paid	approved
`fraud_scenario`	string	Specific fraud scenario label (e.g., staged accident, inflated claim, duplicate claim, false injury)7 values · optional	inflated_claim
`event_features`	string	Serialized JSON string containing detailed event-level features (e.g., sensor data, witness statements, repair estimates)optional	{"impact_speed":26,"polic…
`customer_gender`	string	Gender of the customermale · female · other · · optional	female
`customer_occupation`	string	Occupation of the customeroptional	designer
Numbers 4 columns
`claim_amount`	float	Total monetary amount claimed by the customer0 or more	1750.45
`reported_delay_days`	integer	Number of days between incident date and claim filing date0 or more · optional	2
`number_of_previous_claims`	integer	Total number of previous claims filed by the customer0 or more · optional	0
`customer_age`	integer	Age of the customer at the time of claim0 or more · optional	29
Dates and times 4 columns
`claim_date`	date	Date when the claim was filed	2024-03-05
`incident_date`	date	Date when the incident leading to the claim occurred	2024-03-03
`policy_start_date`	date	Start date of the insurance policyoptional	2022-10-01
`policy_end_date`	date	End date of the insurance policyoptional	2025-09-30
True or false 1 column
`fraud_label`	boolean	Indicates whether the claim is fraudulent (true) or legitimate (false)	false

Use it for

An insurance dashboard
The fraud_label rate, claim_amount by claim_status and a breakdown of incident_type. Excel, Power BI or Tableau.
Why do 58 of 199 rows have fraud_label = true?
A root-cause class exercise
Hand out the rows and one question. The answer is in the data, not in the brief.
Claims200
CLM000011750.45approved
CLM000200.01rejected
CLM000211000000under_re…
A software demo
Believable claims with policy_id, customer_id and claim_date to fill a screen in front of a buyer.

This dataset200 rows23 columns

Yours1,000 rows23 columnsincident_location_street: UK only

blueprint · insurance-claim-fraud-scenarios

Behind this dataset

Same schema. As many rows as you need.

These 200 rows came out of a blueprint — 23 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

Claims include policyholder demographics
Fraud scenarios labeled with type
Claim value and settlement time tracked
Duplicate and exaggerated claims simulated

Rows1,000 rows10,000 rows100,000 rows1,000,000 rows

Open the blueprint in Data Factory

1 credit per row. Preview 20 rows free. Buy rows for the full run.

Exports: CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
Licence: yours to use, including commercially
API slug: insurance-claim-fraud-scenarios

What should your data show?

No signup. No card.

Sample rows

What the 200 rows show

Columns

Use it for

An insurance dashboard

A root-cause class exercise

A software demo

Make it yours.

Same schema. As many rows as you need.

More in Insurance

What should your data show?