Fraud Detection
Jumio needed to verify online identities accurately across thousands of document types and jurisdictions while detecting fraud, deepfakes, and screen attacks that are becoming more advanced as AI capabilities spread.
Reported impact
Jumio supports identity verification for more than 5,000 ID types across over 200 countries and territories and uses over 50 ML models to keep verification friction low for legitimate users while improving fraud defenses against deepfakes and other attacks.
Jumio uses AWS to operate a global identity-verification platform that must distinguish legitimate users from fraud, deepfakes, and screen attacks across many document types and jurisdictions. Amazon SageMaker supports the machine-learning lifecycle for more than 50 models involved in Jumio’s verification pipeline, while Amazon Rekognition provides image-analysis capabilities used in secure, frictionless identity checks. Submitted identity documents and biometric evidence are evaluated through these AI and ML services so Jumio can assess liveness, detect manipulation, and return verification results that support customer onboarding, age checks, and other regulated transactions. Running the stack on AWS gives Jumio a scalable foundation for global data-sovereignty requirements and helps the company continuously adapt its fraud controls as generative AI makes attacks more sophisticated.
Identity documents and user imagery are submitted into Jumio's verification workflow, analyzed by Amazon Rekognition and the ML models managed on Amazon SageMaker, and returned as identity-verification and anti-fraud decisions for customer transactions.
In total, we work with over 5,000 ID types in over 200 countries and territories.
Learn how Jumio is using Amazon SageMaker and Amazon Rekognition to deliver a seamless and secure identity verification experience.
Our goal is to deliver solutions that convert as many legitimate customers as possible while identifying as much fraud as possible.
We have over 50 ML models involved in the IDV process, determining whether the people on-screen are live humans or some type of deepfake or screen attack.
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
The profile says, "Browse Jumio Corporation's employee directory of 617 employees.", which maps to Mid-market.
Related
AWS Sample Sep 16, 2024 Gen When creating Norton Genie, Gen faced several development obstacles. Its team wanted to develop a prototype within weeks instead of months, an incredibly quick turnaround. A study conducted by the Social Market Foundation found that one in five people were victims of scams between… AWS Lambda · AWS Step Functions · Amazon DynamoDB AWS Sample Apr 1, 2024 Solink As a company born in the cloud, Solink provides video analysis and AI-powered analytics to retailers who have historically treated security cameras as an insurance policy to manually retrieve footage after an incident… "We can reduce shrink by up to 70 percent for a company with Solink," says Farrell. Amazon Rekognition · Amazon S3 · Amazon SageMaker AWS Sample Jan 17, 2024 Veriff In today’s connected world, knowing who’s on the other side of a screen isn’t a luxury—it’s a requirement. Every online transaction, every new account sign-up, and every access request hinges on making sure that people… 1 second Amazon Rekognition · Amazon SageMaker 