What is Image Authentication?

The digitalization of the world emphasizes the importance of fraud detection. The need for the insurance industry to be tech-enabled is being widely accepted but not widely applied. We understand how difficult it can be to guard yourself against fraudulent claims. This is why we want to inform you of the best tools to verify claims.

Our industry is one of the few places where believing what you see is a risk and you need to be informed.  Understand what Image Authentication is through its applicable uses and the different tools it can offer you.

What is Image Authentication?

Image Authentication is the process of verifying the integrity of an image or video to ensure that it has not been altered, manipulated, or misrepresented.

An Image goes through a three-step process from start to finish:

The Point of Capture

When a camera is put to use to capture a photo or video it automatically leaves a unique fingerprint at the source. This ‘virtual fingerprint’ encompasses the image’s contents of course, but also sensitive metadata.

Metadata is best defined as the information of an image. This includes specific details like the time, location, and author of the image.

Differentiating The Fingerprints

Most if not all, image tampering operations leave behind ‘virtual fingerprints’ in the form of image distortions. However, since each image is unique and depends on what the fraudster is trying to achieve, they are all very different from each other. The best fingerprint features to trace are resampling, multiple compressions, duplicated regions, and noise inconsistencies.

Resampling:

Used to alter the dimensions of an image and make it smaller or larger. This process changes the pixel values to then adjust the resolution or size of an image. When someone tampers with an image, they may need to resize or resample it to fit it into a different context or to hide evidence. By analyzing these changes you can determine if the image has been altered.

Multiple Compressions:

Every time an image is compressed it loses quality and pieces of evidence. Let’s say you have a photo on your computer. First, you compress it as a JPEG. Then, you take the already compressed image and compress it again into a PNG. Now, you have an image that has undergone multiple compressions. To verify the authenticity, you would reverse the process. If the different stages don’t match this indicates manipulation.

Duplicate Regions:

An identical or similar content piece appears on multiple parts of a photo. Used to cover up something on an image. Let’s look at this scenario, someone copies a patch of grass over the subject on an image. This can make it seem the person is somewhere they are not and derail the evidence.

Noise Inconsistencies:

this refers to irregularities or variations in the background of images. Noise inconsistencies challenge the reliability of the data. In an unaltered photo, the level of blurriness should stay consistent throughout. However, in an image that has been tampered with the level of blur in that area might differ or increase.

Verification of Reality

The image will undergo a meticulous process to will all modifications that have been made, instilling confidence in anyone who sees the evidence. These are some analyses used to verify the authenticity of an image:

File Format Analysis:

Images that are camera originals will match the details of the device used to take the photo. File format analysis determines this.

Global Analysis:

When traces of image resizing or resaving are found then we know for sure this image is not a camera original. But doesn’t confirm it has been altered.

Local Analysis:

This examines the image for local modifications which are done when hiding or adding a subject.

Camera Identification:

Matches the device used to take the image. This is most useful if the device is in custody.

The Uses of Image Authentication

Images are now being treated the same as writing in a court of law. According to the People v Beckley It was decided that “A photograph is a “writing” and “[a]uthentication of a writing is required before it may be received in evidence.” People v. Beckley, 185 Cal.App.4th 509, 514 (Cal. Ct. App. 2010) This emphasizes the importance of authentication especially if it is being used as landmark evidence and there are no witnesses to verify the image.

Any of the following instances requires image authentication:

1. Presenting footage in court
2. Verifying/denying a claim
3. Investigating case evidence

The advancements in photoshopping are tempting for people trying to get out of a claim or verify they deserve a payout. Today we are seeing tampered images of everything from injuries to Donald Trump being arrested.

Shallow fakes

This new term for edited media known as shallow fakes is becoming a real concern for the insurance industry. Shallow fakes create misleading or fake videos used to make fraudulent insurance claims. Few insurance companies have image authentication systems in place resulting in many false claims being approved.

Imagine someone altering a video to make it look like a fake accident or injury occurred. This raises premiums for everyone even those not involved in these schemes. The challenge for insurers results in financial losses from paying out fake accidents.

What can you do?

Bringing in a team of highly experienced video forensics experts who can analyze, compare, and evaluate anything from simple file conversion to the metadata of an image is priceless in claims verification. The application of evidence rules from the pre-digital era is no longer working. The authentication of a photograph in court is now a minimum requirement. As technology advances the need for verifying manipulated media is in high demand. Particularly with sectors like insurance where the ramifications of fraudulent claims are significant.

Provide your team with expert resources. Contact our digital investigations department today.