How blockchain could help block fake news
Creating an immutable ledger to track metadata associated with published content shows promise as a method for identifying authentic news
18 February 2020 | 0
In 2018, a video of former President Barrack Obama surfaced on YouTube explaining how easily technology could be used to manipulate video and create fake news. It got more than 7.2 million views.
In the video, Obama explains how we live in dangerous times when “enemies” can make anyone say anything at any point in time. Moments later, it is revealed that the video was itself faked.
Whether its news articles, images or video, fake and misleading content has proliferated across the Internet over the past five or so years. One possible solution to the problem now being proposed would standardise how content is delivered online, with anything outside those standards not trusted.
Enter blockchain as a method of whitelisting news and other Web content.
As part of a Web 3.0 evolution, blockchain is being used to create a decentralised Web, where an immutable ledger records information about content and then is inextricably linked back to that content to ensure its authenticity. Personal data stores would also allow companies and individuals to retain control over content they make or consume.
By 2023, up to 30% of news and video content world-wide could be authenticated as real by blockchain – in effect, countering deep fake technology, according to the Gartner 2020 Predicts report released in December.
“Blockchain can track the provenance of news (text or video content) so that consumers of the content know where it came from and are assured it has not been altered,” Avivah Litan, a Gartner vice president of research, said in a recent blog. “Putting social media and social networks on blockchain will enable users to control not only their own information, [but also] the algorithms and filters that direct their information flows.”
In December, Twitter CEO Jack Dorsey announced that Twitter is “funding a small independent team of up to five open source architects, engineers, and designers to develop an open and decentralised standard for social media”.
Twitter’s concept is to allow users to read what they want, offering filters that could be based on blockchain to authenticate content if they want or unfiltered content if they chose that, according to Litan.
Over the past two years, several high-profile projects using blockchain aimed at thwarting fake news have emerged.
The New York Times-sponsored News Provenance Project, the Deep Trust Alliance and PO.ET, are all examples of efforts to standardise the way news, images and video is pushed out to the Web, while recording the journey it takes from source to consumer.
Po.et is developing a decentralised system based on Bitcoin blockchain as an immutable record that timestamps content and uses current media industry interoperability standards.
The News Provenance Project worked with IBM’s Garage to develop a proof of concept using Hyperledger Fabric blockchain to store contextual metadata about news photos and video, including who shot footage, where it was shot and when it was edited and published. The blockchain would record a photo’s origins: when, where and by whom it was taken, who published it and how it had been used across a network of news organisations.
The Deep Trust Alliance was founded by Kathryn Harrison, who was formerly director of global product management for IBM’s blockchain effort, where she contributed to Hyperledger Fabric development and building the company’s managed service on top of it.
“Most companies and efforts start with deep fake video and images, which aren’t even prevalent on the Internet; that’s such a small percent of fake content,” Litan said. “They started with deep fakes because they’re the easiest.”
Unlike manually manipulated content, such as a news article created or edited by bad actors, deep fake news refers to an image or video that is generated by AI and machine learning technology known as generative adversarial networks (GANs).
Danny O’Brien, director of strategy for the Electronic Frontier Foundation, said the use of technology to solve a social problem rarely works.
“I’d happily bet 10 bitcoins 30% of news not authenticated by blockchain by 2023,” O’Brien said, referring to the Gartner prediction. “Some people believe that the solution to the problem with fake news is to create a system to vouch for true news. That stumbles in a couple places. First, just because you put it out doesn’t mean it’s true. And secondly, people responding to and [sharing] fake news are rarely motivated to find out if it’s true.”
In other words, consumers are not typically interested in using critical thinking to determine whether news is real or not – as long as it fits a narrative they like.
The New York Times’ News Provenance Project, however, did not begin with articles. It started by experimenting with images, because it is easier for the software to examine pixels and determine whether they had been changed or remained authentic, Litan said.
In contrast, the best method for authenticating written text would be an industry consensus around the sources that produced it. In other words, using blockchain’s native consensus algorithm to allow content producers to agree when something is authentic before it can be published.
“The best use for blockchain is content management. If everyone adopted a content management system that [cryptographically] signed your stories and every edit made to them, and then recorded that and authenticated it using blockchain, the chances of it being fake would be zero,” Litan said.
There is, however, malicious fake news created by state operators who construct narratives for political purposes. It matters little whether they join an authentication system because any such system will never be universal, O’Brien said.
“If you try to throw your weight behind a system that creates a badge of authenticity for ‘legitimate’ news, either that won’t work for getting rid of it or you can’t get everyone to sign up,” O’Brien said.
Litan agreed, saying there is no chance all media outlets will agree on a single content management system.
Along with being used to white-list authentic news sources, however, Litan said blockchain is being explored as a method for:
- Authentication: Other technologies such as electronic impulse testing or spectral imaging to ensure the authenticity of the article (eg news, food, electronics) being recorded and tracked on the blockchain
- Blacklisting or anomaly detection: using AI and machine learning models to detect behavioral or data anomalies for assets/information/activities both on and off the blockchain that feed into the blockchain
- Bridging physical and virtual truth: IoT sensors/networks/digital twins to connect physical ‘truth’ to virtual business event ‘truth’ recorded on the blockchain
“A decentralised social network platform could still feed you distorted sensational news, but presumably only if that’s what you choose to read. With blockchain provenance you could also be assured for the source for that news,” Litan said. “In the end, it will be your purposeful choice to read fake news from bad actors.”
IDG News Service