Podium Facial Validation

Unproctored assessment verification

Podium Facial Validation technology allows you to validate your remote unproctored assessments. Facial Validation uses the latest web technologies coupled with sophisticated Artificial Intelligence (AI) to monitor an on-going assessment and record activities that may invalidate the results.

In accordance with guidance from The International Privacy Subcommittee of the Association of Test Publishers (ATP) Security Committee, Podium has developed its Facial Validation to detect a variety of potential activities a candidate may attempt while completing a remote assessment, especially one that is of high stakes.

The Facial Validation system can detect who is completing the assessment, whether that person is distracted while completing the assessment, whether they receive aid from a friend or colleague or whether the candidate is different at any point of the assessment. A response profile is then generated using AI facial matching technology, and a report is presented for the administrator to validate the assessment process.

FAQ: Is this using Facial Recognition?

No. Facial Recognition is a very different technology, one that is generally seen as containing a high degree of bias, especially towards certain races. Facial Recognition taps into large, specialised security databases and uses AI to match someone’s face. This automated process has led to many misidentifications and is generally considered unethical to use.

Facial Validation is very different. It takes a reference image before the assessment, freely offered by the informed candidate. During the assessment process it watches the candidates face and activity. The number of people can then be identified, and a match between the reference and the live images can be made. A human is then given a portfolio of evidence to help them make a decision.

FAQ: Is the computer making the decisions for me?

No. Facial Validation is a passive approach to validating your remotely proctored assessments. It watches and records potential activities performed by a candidate. These recordings are then scanned using the AI to identify the type of behaviour identified. A report portfolio is then created, either for an individual or as a group of candidates and made available to the human administrator.

This portfolio lists the type of behaviours observed, and the degree of accuracy the AI has matched the behaviour. The human administrator can then use this portfolio of evidence together with any other data they may already have to form a decision.

FAQ: What happens to the images once they’ve been taken?

All Facial Validation images are encrypted, so that may not be viewed, and stored in a secure data centre. The Podium AI scans these images and scores each one. A portfolio of evidence is then generated for the administrator to review. Some, but not all, of the images are chosen, especially if the AI has identified specific behaviours, for the portfolio. No other application or service has access to this secure location.

The stored images are later permanently deleted in accordance with the Podium Data Protection & Privacy policies. In accordance with GDPR a candidate is entitled to request data deletion if they wish.