This weekend, an image circulated on Twitter of huge crowds in London calling for the far-right figure Tommy Robinson to be released from prison.
Except the image wasn’t of crowds in London and they certainly weren’t protesting for the English Defence League leader to be released.
Instead, it was a 13-year old picture of football fans in Liverpool celebrating the club’s Champion League victory.
That’s why a Belgian media company, Look Live Media, is launching VeriFlix, an AI-powered software which can source and authenticate user-generated footage so journalists can differentiate between real and fake news online.
How VeriFlix AI technology works
The technology uses AI to help journalists verify submitted footage. The idea is that news organisations can put out requests for video, and then use VeriFlix to ensure it is real footage.
For instance, says a news organisation requests video from a protest in central London. People within the specified geofenced area of central London will be able to film their surroundings and submit the video for VeriFlix authentication.
The software then uses AI to authenticate the meta data of submissions. It checks the geolocation, time stamps and footage quality.
As well, the tech also considers a video’s relevance, consistency and diversity, to ensure that each submission is a valid reflection of the event or story.
Donald Staar, founder and CEO of Look Live Media, said: “As fake news continues to be an increasingly difficult and recurring problem for media companies to tackle, validating and authenticating user-generated content to ensure accurate representation of an event or news story is our main priority.”
At the moment, VeriFlix is available via an app or a software development kit, which can be integrated into a news organisation’s existing app.
Fighting fake news with technology
There are several initiatives out there to tackle the spread of fake news. For instance, there’s the fun Bad News game which aims to be a “vaccination” against the disease of fake news.
What’s interesting for the VeriFlix project is that it has been funded through Google’s Digital News Initiative, a fund which supports innovative high-quality digital journalism.
The project received funding through a Belgian broadcasting company, Roularta Media Group. It’s through Roularta that VeriFlix was awarded €400,000 (£353,000) through the Google initiative.
Thanks to this funding, Look Live Media was able to develop VeriFlix in partnership with an AI team at the Belgium university, KU Leuven.
Staar added: “Having Google DNI’s backing only reinforces our guiding belief: that Veriflix can accurately and successfully solve this growing issue for media organisations and their audiences across the world, improving trust, transparency and quality reporting in the process.”
- AI is still terrible at spotting fake news
- Watch These Activists Explain How Technology Helped Them Fight Oppressive Governments
- Europe eyeing bot IDs, ad transparency and blockchain to fight fakes
- The fake video era of US politics has arrived on Twitter
- OpenAI let us try its state-of-the-art NLP text generator
- Tech platforms should fight Islamophobia the way they fought ISIS
- Hobbs & Shaw is joyously stupid about science and technology
- Full transcript: Data for Democracy policy head Renée DiResta answers disinformation questions on Too Embarrassed to Ask
- If you can quit social media, but don’t, then you’re part of the problem, Jaron Lanier says
- Who needs democracy when you have data?