Vloggi adds video provenance layer for human clips

Vloggi has introduced a video verification layer to its submission platform, linking each contribution to its source and submission history.

The Sydney-based company says the update is intended to help organisations establish the provenance of human-generated video as concern grows over deepfakes and other synthetic media.

The system creates a documented chain of custody for each video submission. That record can include the contributor's identity, the permissions attached to the footage, whether the clip was captured through Vloggi or uploaded from an existing file, and signals related to editing or recompression.

It also embeds the contributor's name in the file's author metadata, allowing attribution information to remain attached to the video after it leaves the platform. Contributors can download a unique submission ID to help track how their footage is used.

Election focus

The rollout comes as Vloggi looks to meet demand from political campaigns in the United States ahead of the midterm elections. It has also been in discussions with campaign teams in the UK and Australia.

Previous users of the platform include the presidential campaigns of Robert F. Kennedy Jr. and Joe Biden, which used it to gather voter sentiment through qualitative video surveys. Former UK Prime Minister Rishi Sunak has also used the technology to collect supporters' views.

Justin Wastnage, Founder and Chief Executive Officer at Vloggi, linked the update to the wider rise in synthetic media.

"The move effectively acts as a barrier against nefarious actors using generative AI to create deep fake content in business, marketing or politics," said Justin Wastnage, Founder and Chief Executive Officer at Vloggi.

Concern over the authenticity of digital evidence and online media has spread beyond election campaigns. Courts, newsrooms, researchers and companies are all grappling with how to verify whether video footage is genuine and who created it.

Vloggi pointed to recent legal cases in the United States that highlight those concerns. In one California housing dispute, a judge found that a witness video presented as genuine had been generated using artificial intelligence. In another case in New York, child-abuse findings were overturned because video evidence had not been adequately authenticated.

"As we head into a new election cycle in the US, it's more important than ever to reliably determine whether video content has been created by actual humans or machines," said Wastnage.

Wider use cases

The verification layer is also intended for use in corporate communications, employee feedback, market research, journalism and education. Among Vloggi's existing customers is one of Australia's largest telecommunications providers, which uses video forms to gather sensitive employee feedback.

According to Vloggi, a pilot project with a major US bank is also planned, with the bank seeking a way to collect video from identified human contributors without synthetic submissions entering its workflow.

Wastnage said the issue is not only detecting fake material, but also proving the origin of real footage.

"We see entire brands' outputs of supposed 'user-generated content' created programmatically without a single human having been filmed," said Wastnage.

"While the new wave of AI slop can be useful for brands, we see it as dangerous for areas like politics, research and education, where authenticity of the video is paramount."

Vloggi says its approach addresses a gap left by Content Credentials, which major technology groups have promoted to identify synthetic or AI-created material. Its system is designed to provide an equivalent record for video made by identifiable people, backed by consent and submission data.

At the centre of that approach is a revised consent framework. Vloggi says it has clarified who receives re-use rights for contributed footage and on what basis, in an effort to tighten accountability around how user-submitted video is handled after collection.

"As synthetic video becomes easier to produce, and authentic footage increasingly difficult to prove, this upgrade creates an immutable record showing who submitted a video, when it was received, what permissions were granted and how the footage entered an organisation's workflow," said Wastnage.

"It is essentially a chain of custody for human-generated video."

Privacy changes

The verification update follows broader changes to the platform's contributor experience. Vloggi has added interfaces in French, Spanish, Portuguese, German, Dutch and Italian, and updated its privacy framework to support requirements under the European Union's GDPR and California's CCPA.

Those changes include clearer legal disclosures, translated privacy information and distinctions between data-controller and data-processor responsibilities. The work is intended to make it easier for organisations to gather attributable video submissions across Europe and the Americas.

Wastnage said Vloggi does not see human-verified and AI-generated video as mutually exclusive. He said the company uses artificial intelligence in parts of its own business, including product development, image generation and drafting copy, but draws a distinction when a verifiable human source is required.

"The challenge is no longer simply detecting fake video," said Wastnage.

"Organisations also need a practical way to establish where real video came from, who submitted it and what permission they gave for it to be used.

"If AI-generated content is going to carry a record of where it came from, authentic human content needs an equally serious record of its origins," said Wastnage.

"This is about protecting human attribution and consent. Real people should not become anonymous raw material for organisations or AI systems."