Artificial intelligence (AI) is the broad concept of machines being able to carry out tasks that, if performed by a human, would require the person to use their “intelligence” to perform the tasks.
Machine learning is a popular application of AI where machines are fed large amounts of data and learn to make predictions on new data.
Augmented intelligence emphasizes the fact that AI technologies have been developed specifically to help humans, rather than to replace them.
In this way, augmented intelligence applications combine human and machine intelligence. It is important in systems where the risk of failure is too great or the AI is not evolved enough to take humans completely out of the equation.
Some identity verification solution providers rely solely on artificial intelligence to perform ID and “selfie” checks for online identity verification, claiming it eliminates the need for human review. The truth is, AI is almost, but not entirely ready to remove humans from the identity verification process.
Without augmented intelligence (via expert, real-time human review) your customers are frequently redirected to manual processes for verifications that fall in the “maybe” zone, which can result in:
Need for Big Data
Effective algorithms require huge datasets for each type of ID you wish you verify.
Number of ID Types and Subtypes
A provider like Jumio can support over 3,500 constantly-changing ID types and subtypes. AI models must support them all.
Blurriness, Bad Lighting and Glare
Picture quality is dependent on user’s ability to capture a good image from their device.
The Challenge of Omnichannel
Verifying customers across devices presents additional challenges for AI models.
The Selfie Requirement
Many online companies require a selfie to ensure the picture in the selfie matches the picture in the government-issued ID. AI models are still perfecting this “match,” especially when ID photos are outdated.
New types of fraud, including spoofing attacks on selfies, can stump an AI model.