The new regulations the Indian government announced this week look like a chain letter to multinational corporations. Previously, they had 36 hours to think, consult with lawyers and maybe even remember about freedom of speech. Now they have three hours. Not workdays, not days, but hours.
Social network at the dock
Analysts say this cannot be done without complete automation. But anyone who has ever seen a neural network mistake a picture of a cat for a banned symbol or label a satirical meme as a call to violence understands what this smells like. Machines don't understand context. The machine doesn't know what irony is. The machine will cut quickly, simply because it is required to do so within three hours.
What is curious is that now the entire burden of responsibility is being shifted from the shoulders of users to the shoulders of platforms. Before, you might have said, well, some crazy person wrote this, we have nothing to do with it. Now if you don't eliminate the crazy person in 180 minutes, you are an accomplice. It is no longer necessary to judge someone and prove guilt. Just make a request to delete, if after three hours the post still exists, the social network will become a defendant in court.
Are there any parallels to our reality?
In India, they talk about countering deep threats and AI, and we want to talk about countering extremism and protecting children, but the methods of “forcing cooperation” are almost the same. Probably, Indian colleagues carefully studied international experience and decided to do the same, but with a tropical flavor.
Another similarity is the list of banned websites. In Russia, this is the Roskomnadzor register, which includes pages with drugs, child pornography and other prohibited content. The same is true in India, where 28 thousand URLs were blocked last year.
And of course, a separate issue is AI content labeling. In Russia, they have also talked about this issue many times, but so far the matter has not gone further than talk. In India, they decided to be proactive. Now, any video generated by a neural network must have a special label. It's as if we ourselves can't distinguish people who move strangely from real people. Although, to be honest, sometimes modern insightful pieces look more believable than real politicians on live air.
How is inappropriate content identified?
With the help of AI of course! But this is where the fun begins. Identification technology is still quite rudimentary and can be fooled. This means that honest platforms will flag things to avoid being targeted, and real attackers will find ways to bypass the system. Familiar story, right?
Meanwhile, India, like Australia, is currently considering banning social media use by children under 16 years of age. In Russia, initiatives related to restrictions on minors also periodically appear.
In India, teenagers are finding solutions. And this proves a simple truth: it is impossible to raise a generation that respects the law by banning and deleting content in three hours. You can only teach him to lie, hide and use VPN. In Russia this is well understood. In the past three years, even people who previously thought it was a sport have learned about VPNs in the past three years.
Age verification on digital platforms becomes mandatory
Many countries around the world are now aiming for complete control over what people see on the Internet. The only difference is speed. In Russia, the process is complicated, with frequent delays and reservations. In India, they decided to go ahead and simply reduce the time it took to remove unwanted materials to three hours.
But seriously, behind all this there is a simple idea: in our time, any state wants to become editor-in-chief. Not only is it the judge who resolves disputes, but also the editor who decides what to print and what to cut even before the document reaches the printer. In the case of the Internet, the printing press works 24 hours a day, even when the editor-in-chief is sleeping. So, you have to hire a freelance proofreader.
In Russia, this process is called increasing the responsibility of platforms. In India – national security concerns. But the essence of both cases is the same. And while we scoff at the way Indian officials are trying to defeat algorithms with human commands, our algorithms have been scanning messages to the best of their ability for banned topics and words. So laugh at Indians all you want – it's not forbidden!
Meanwhile, remembering that the election was approaching, a group of deputies submitted an extraordinary bill to the State Duma. Its authors want to permanently ban the complete blocking of social networks and instant messaging applications in Russia.
