In an era of spam bots and artificial intelligence deep fakes, it’s becoming increasingly difficult to know the truth. It’s one of the biggest challenges that modern social media companies face, requiring cutting-edge technology to combat a never-ending wave of content published by bad actors.
Thankfully, artificial intelligence and machine learning can provide efficient and scalable solutions to counter the spread of the very same problem they’re often used to create. These tools are utilized by experts like Vineet Dhanawat, a software engineer and machine learning expert who is solving this crisis for the biggest tech companies in the world.
Learn more about how Dhanawat is creating tools that reduce harmful and misleading content while penalizing those who spread it — combating spam at scale and helping social media rebuild trust among users.
Identifying and Reducing Potentially Harmful Content
Social media abuse is becoming an increasingly significant problem. In 2024, 36% of social media users reported seeing false or misleading political content, and less than half of users trusted social media for general news and information. Independent fact-checkers go a long way toward combating this issue, but it’s a losing battle — with billions of images shared on social media every single day, the task of identifying spam is too large and complex for human fact-checkers alone.
Enter Dhanawat and his expertise in machine learning. Building upon his experience at industry giants like Oracle and IBM, he’s developed neural net-based models for recognizing near duplicates of images — creating matching algorithms that can automatically identify and label images that are similar to those that have been identified as false by human fact-checkers. Even if an image has been altered to look like new content, Dhanawat’s algorithm is able to flag it with a fair amount of accuracy.
This tool can process countless images in mere seconds and compare them to a database of red-flagged content, allowing for instant identification and immediate removal. By automating this process using machine learning, Dhanawat has helped some of the biggest names in tech accomplish what would otherwise be impossible.
But Dhanawat’s work has also expanded into the theoretical, driving machine learning innovation at scale as he tackles the big picture — he’s successfully filed various patents and published research papers in top journals, becoming a leading voice in the industry as he continues to fight spam behind the scenes.
Safer Systems for More Transparent Platforms
But abuse doesn’t stop just because you remove the offending content. You have to have a way to identify offending users and prevent them just trying again later. To this end, Dhanawat single-handedly built the penalty system for a Fortune 500 social media platform, establishing firm consequences for users who repeatedly post spam. His penalty system limits the ability of malicious accounts to advertise, removes their posts from suggestion algorithms, and generates pop-up messages informing users of pages that repeatedly post harmful content.
By reducing the distribution of fake content, removing offender accounts, and preventing these same users from registering with other accounts, Dhanawat’s impact on the content crisis is clear, with the direct result of these policies being a reduction of hoax and spam accounts by 10% by one of the biggest social media platforms in the world.
In 2019, this Fortune 500 company discovered 2.2 billion fake accounts and 2.9 billion instances of spam on its platform. But following Dhanawat’s efforts, the future is already looking brighter: Only 631 million fake accounts were flagged in the first quarter of 2024, while 436 million spam posts were removed during that same period.
Real World Problems, Machine Learning Solutions
Vineet Dhanawat aims to continue to use machine learning to make safer, more transparent social media platforms, thereby winning back the trust of billions. His work in building better tools for fact-checkers and making their methods more transparent is going a long way toward alleviating the tangible harm caused by spam and harmful content and making the digital ecosystem safer for users worldwide.
Though the task of combating social media abuse is only getting harder, Dhanawat is committed to making sure we can trust the platforms that connect us to the rest of the world.
