Microsoft rolls out tools to spot disinformation, deepfakes

The tools will help organisations, news outlets and political campaigns fight disinformation   | Photo Credit: Reuters

Microsoft announced on Tuesday tools to identify manipulated videos and photos. The tools are part of the company’s programme to help protect voting through ElectionGuard, a feature to verify and secure campaigns ahead of US elections, Microsoft said in a statement.

Microsoft’s Video Authenticator can analyse a video or a still photo to provide a percentage chance, or confidence score, that the media is artificially manipulated. In the case of a video, it can provide the details in real-time on each frame as the video plays.

It works by detecting the blending boundary of the deepfake and subtle fading elements that might not be detectable by the human eye.

The tool was created taking information from a public dataset and testing it on a deepfake detection dataset. Microsoft built the tool in partnership with the AI Foundation based in San Francisco, to make it available to organisations, news outlets and political campaigns.

To combat disinformation, the technology company has added a tool into its cloud service Azure. The tool enables content producers to add digital hashes and certificates that are attached as metadata and carried along with the content.

Microsoft has also added a reader that checks the certificates and matches hashes, letting people know with a high degree of accuracy that the content is authentic and that it hasn’t been changed. It also shares details about content creator.

The new features will also allow users to identify fake content outside of election campaigns, the statement said.

This article is closed for comments.
Please Email the Editor

Printable version | Mar 3, 2021 7:34:03 AM |

Next Story