Edward Greenspon and I have an oped on the likely evolution of #fakenews: a pernicious mix of AI, commercial surveillance, adtech and social platforms that is going to undermine democracy in some critical ways. If fake text had a social impact on our understanding of events and news, imagine what the coming fake and micro-targeted video and audio is going to do. Watch this space, it’s going to get wild fast.
‘Fake news 2.0’: A threat to Canada’s democracy
Ed Greenspon and Taylor Owen, The Globe and Mail, May 29, 2017.
The muggings of liberal democracies over the past year by election hackers and purveyors of fake news are on the cusp of becoming far worse.
By Canada’s next federal election, a combination of artificial intelligence software and data analytics built on vast consumer surveillance will allow depictions of events and statements to be instantly and automatically tailored, manipulated and manufactured to the predispositions of tiny subsets of the population. Fact or fabrication may be almost impossible to sort out.
“Fake news 2.0” will further disorient and disillusion populations and undermine free and fair elections. If these were physical attacks on polling stations or election workers, authorities would respond forcefully. The same zero tolerance is required of the propagation and targeting of falsehoods for commercial, partisan or geopolitical purposes. The challenge is that unlike illegal voting, which is a clearly criminal act, the dissemination of misinformation is embedded in the very financial model of digital media.
This is serious stuff. Germany is looking to hold social media companies to account for false content on their sites. Britain’s Information Commissioner is investigating the political use of social media-generated data, including the activities of an obscure Canadian analytics firm that received millions from the Leave side in the Brexit campaign. In the United States, investigative reporters, foundations and academics are unearthing startling insights into how the dark side of the digital ecosystem operates.
Fake news is inexpensive to produce (unlike real news); makes strange political bedfellows of the likes of white supremacists, human rights activists, foreign powers and anti-social billionaires; and plays to the clickbait tendencies of digital platforms. A recent study, The Platform Press: How Silicon Valley reengineered journalism, argues that the incentives of the new system favour the shareable over the informative and the sensational over the substantial. Fake news that circulated during the 2016 U.S. election is not a one-off problem, but rather a canary in a coal mine for a structural problem in our information ecosystem. On platforms driven by surveillance and targeted advertising, serious journalism is generally downgraded while fake news rises alongside gossip, entertainment and content shared from family and friends.
As with classic propaganda, fake news seeks credibility via constant repetition and amplification, supplied by a network of paid trolls, bots and proxy sites. The core openness of the Web enables congregations of the disaffected to discover one another and be recruited by the forces of division – Breitbart News, ISIS or Vladimir Putin.
The classic liberal defence of truth and falsehood grappling, with the better prevailing, is undercut by filter bubbles and echo chambers. It has become almost impossible to talk to all of the people even some of the time.
And so the polluted tributaries of disinformation pouring into the Internet raise a critical governance challenge for open societies such as Canada: Who will speak for the public interest and democratic good in the highly influential, but privately owned, digital civic space? What does it mean for a handful of platform companies to exercise unprecedented control over audience and data? How does government clean up the pollution without risking free speech?
Canada needs to catch up on analyzing and responding to these new challenges. Here’s where we would start:
- A well-funded and ongoing research program to keep tabs on the evolving networks and methods of anti-democratic forces, including their use of new technologies. Government support for artificial intelligence is necessary; so is vigilance about how it is applied and governed.
- Upgraded reconnaissance and defences to detect and respond to attacks in the early stages, as with the European Union’s East StratCom Task Force. Prime Minister Justin Trudeau has already instructed his Minister of Democratic Institutions to help political parties protect against hackers. That’s good, but a total rethink of electoral integrity is required, including tightening political spending limits outside writ periods and appointing a digital-savvy chief electoral officer.
- Measures to ensure the vitality of genuine news reporting; fake news cannot be allowed vacant space in which to flourish.
- Transparency and accountability around algorithms and personal data. Recent European initiatives would require platform companies to keep data stored within the national boundaries where it was collected and empower individuals to view what’s collected on them.
Finally, the best safeguard against incursions on commonweal is a truly inclusive democracy, meaning tireless promotion of economic opportunity and social empathy. As Brave New World author Aldous Huxley commented in 1936, propaganda preys on pre-existing grievances. “The propagandist is the man who canalizes an already existing stream. In a land where there is no water, he digs in vain.”