Fake news, deepfakes & real money – a terrifying scenario

Twitter has removed nearly 1,000 accounts and suspended 200,000 others linked to a campaign by the Chinese government against protesters in Hong Kong. A couple of hundred Twitter and Facebook accounts spreading fake news over the Indian government’s decision to remove a special legislation for the Himalayan state of Jammu & Kashmir were also shut down by these two companies.

Technological advancements are being used for terrifying purposes that threaten, individuals, organizations and even countries. The use of deep learning has created programs that can mimic the way we speak, write and interact with amazing accuracy which if misused can unleash havoc in our lives. OpenAI, a research organization in artificial general intelligence, got so frightened of its own program which creates amazingly convincing content based on inputs of a few words that it stopped the release of this code.

The Macedonian town of Veles has become the epicenter for fake news. They came into spotlight when misinformation campaigns generating from this town influenced the US Presidential election in 2016. Fake news writers in Veles claim to have earned up to $2,500 a day from advertising on their website; the average monthly income in Macedonia is just $426. The profits come primarily from ad services such as Google’s AdSense, which place targeted advertisements around the web. Each click sends a little bit of cash back to the content creator. Their sight is firmly set on the 2020 US presidential elections.

Equally dangerous are deepfakes. These are Artificial intelligence generated unbelievably realistic videos which are almost impossible to distinguish from the real ones. MIT has warned that deepfake attacks would increase in 2019. Programs are available whereby videos of people can be altered to say anything you want them to say.

Researchers at Harvard University and MIT-IBM Watson AI Lab have created a tool to help combat this menace. Called GLTR (for Giant Language Model Test Room), uses AI to detect the very statistical text patterns that have been used to create fake news. Telecom company Orange and startup Block Expert have launched Safe.press, a blockchain-based consortium for certifying digital news.

As battle between fake news, alternative facts, deep fakes and technologies to counter those rages on, there is hardly any human activity that is not vulnerable to these criminals. The chilling fact is that misinformation is now weaponized by state actors creating a new battlefront, instead of armies occupying territory, cyber-attacks, fake news and deep fakes can cripple a nation’s economic infrastructure in a matter of nanoseconds.