Russian authorities are turning to artificial intelligence tools to monitor internet activity as Moscow tightens its control over what happens online
By Umberto Bacchi
TBILISI, Nov 30 (Thomson Reuters Foundation) - A woman posing in a thong outside a church; a single mother who berated Russian lawmakers and President Vladimir Putin; a saxophonist who criticised World War Two commemorations.
They are among thousands of Russians who have faced court over their social media posts in the past year - a number digital rights groups say could soon turn into a deluge as authorities use artificial intelligence (AI) to police the web.
"We expect that all content posted on social media (in Russia) will be monitored by automated software," said Sarkis Darbinyan, head of the legal department for digital rights group Roskomsvoboda.
"This will be particularly bad for young people who will have red flags put on (them), and be prosecuted and fined for posting different material," he told the Thomson Reuters Foundation by phone.
Russia has passed an array of legislation in recent years to boost what it calls its internet "sovereignty" and tighten control over cyberspace.
The increasing scrutiny of what people say online is part of a wider drive that has seen Moscow pressuring foreign tech firms such as Twitter and Facebook to delete content it deems illegal and blocking opposition sites and news outlets.
While the search for banned material used to be down to police or pro-government activists, authorities are turning to AI tools to quickly wade through millions of posts per day.
Authorities say the monitoring systems are meant to tackle crime, but rights groups fear that they will be used to stifle dissent and crack down on free speech.
"We are seeing a pincer attack by governments across the world with draconian laws that attack freedom of expression and privacy online," said Likhita Banerji, a technology and human rights researcher at Amnesty International.
State communications regulator Roskomnadzor did not reply to a request for comment.
Since 2017, the Perm-based tech firm SEUSLAB has been supplying law enforcement in dozens of regions with software that its director Evgeny Rabchevsky said can process a billion social networking pages and instant messaging chats a day.
He said police use the tool to detect and prevent crimes including terrorism, child pornography, drug-related offences and "destructive subcultures" - a term he said referred to issues like "child suicide propaganda" and calls for violence.
"Authorities use the product to assess social tensions, identify problematic issues of interest ... (and) adjust their activities," Rabchevsky said, adding the firm recently developed an AI tool to monitor social media activity during protests. Last month, the Center for the Study and Network Monitoring of the Youth Environment, an NGO founded on behalf of Putin, said it had developed an AI tool to scan social media for what it considered socially dangerous and destructive content.
The tool was created under a scheme run by youth affairs agency Rosmolodyozh, which did not respond to requests for comment.
Instead, the spokesperson pointed to an interview in Forbes Russia magazine in which the head of the NGO, Denis Zavarzin, said its system was "constantly monitoring" about 1.5 million accounts.
Research shows more Russian AI cyber monitors are in the making.
Official documents seen by Reuters in September show that authorities are developing a new monitoring system that will automatically search for banned content on social networks and the messaging app Telegram.
And tenders are also planned for two more tools, one to search for visual information and the other to defend against information threats.
All three systems are expected to be up and running next year, as draft budget proposals released in September showed Russia may spend 31 billion roubles ($416 million) on enhancing its internet infrastructure in 2022-24.
In Russia, there is no shortage of laws netizens can fall foul of.
In 2019, the country imposed new fines of up to 100,000 roubles on people who spread what the authorities regard as fake news or show "blatant disrespect" for the state.
Court documents seen by the Thomson Reuters Foundation show SEUSLAB's tool was used that year to bring extremism charges against a woman for a blasphemous social media post, in which she said there was no "pederast god".
Activists and others have landed in hot water for posts related to what the government considers "extremist" organisations, including the Jehovah's Witnesses religious group and groups linked to jailed Kremlin critic Alexei Navalny.
Damir Gainutdinov, a lawyer who heads the Net Freedoms Project run by human rights group Agora, said last year journalists and bloggers received more than 1,000 fines for online speech violations.
Gainutdinov believes the fines came from an automated system.
"Even the text of the notice is always the same, it's copied and pasted over and over," he said.
According to Agora, more than 22,000 administrative cases have been initiated since 2017, which include displaying prohibited symbols and disseminating extremist material, with cases nearing a record 7,000 in 2021 alone.
Andrey Shabanov, a saxophonist from the city of Samara, is facing charges including rehabilitating Nazism and disrespecting Russia's military for a series of posts criticising celebrations of the Soviet victory in World War Two.
In May last year, he allegedly decried Soviet totalitarianism on VKontakte, a Russian social media site, and said an annual parade in which people march with portraits of relatives who fought in the war was "idiotic".
The 40-year-old musician also uploaded a photo of Adolf Hitler to a website dedicated to the parade, in a move his lawyer Aleksei Lapuzin said aimed to draw attention to what Shabanov saw as "the growth of fascism in Russia".
Lapuzin, whose client faces a maximum of three years in jail, said the case was emblematic of the shrinking space for online free speech in the country.
Darbinyan of Roskomsvoboda said the deployment of AI vigilantes was all the more concerning with no adequate legal framework for digital rights in place and the government bent on "clearing Russian cyberspace of all undesirable content".
Earlier this year, Putin called for the internet to be bound by moral rules to stop society fraying and railed against what he said was its role in drawing children into opposition street protests, prostitution and drugs.
Shabanov, the saxophonist, said he refrained from posting on social media for a while after his case was initiated, but has now returned to his usual online habits, even though the case is still being heard.
"I don't think words should be something you should be prosecuted for," he said via WhatsApp. "Words or stupid actions are not a crime."
Others are more cautious.
Some rights activists have urged internet users to delete old posts - or stay off social media altogether.
"We do not recommend our supporters use Russian social networks in general. We do not consider them safe," said Mikhail Klimarev, director of the Internet Protection Society, a privacy group.
"We are in a cyber war, with many people in prison and others persecuted for their words on the internet. Self-censorship is becoming more and more common practice. I notice for myself that sometimes I prefer to remain silent." ($1 = 74.5560 roubles)
Can artificial intelligence help close gender gaps at work?
AI surveillance takes U.S. prisons by storm
Digital world seen moving into 'authoritarian space'
(Reporting by Umberto Bacchi @UmbertoBacchi, Editing by Jumana Farouky and Zoe Tabary. Please credit the Thomson Reuters Foundation, the charitable arm of Thomson Reuters, that covers the lives of people around the world who struggle to live freely or fairly. Visit http://news.trust.org)
Our Standards: The Thomson Reuters Trust Principles.