Social Media AI Aims to Foil OSINT Teams

The Artificial Intelligence (AI) inside Facebook, Twitter, and Instagram has fundamentally changed focus in 2021. The good news for insurance is that the new AI has shifted the content users see back to its roots: selfies, personal stories, family photos, documentation of fitness goal achievement and lots of data about peoples’ personal lives.

The creepier news is that refinement of the AI presents a “Stepford Wife” user experience where bragging and lifestyle photos are now center stage. The AI is so laser focused on presenting positive, nonpolitical content that individual users’ data is more exposed than it has been in years. All the political camouflage for personal data (diluting personal photos inside of a long stream of political posts) is stripped away, leaving pure personal information.

So how is social media protecting this user data? The venues are attacking external bots with renewed vengeance. Content bots, scanning bots, scraping bots are all now facing new, complex technical challenges that have been applied across almost every social media platform. The anti bot technology is so aggressive that even human users are being penalized for using ‘bot like’ behavior.

There are consequences for Insurance OSINT teams. As Investigators search social media profiles, their behavior is similar to a bot, where many profiles are opened in rapid succession and lots of friend profiles are also opened. Once flagged as a bot, even human OSINT investigators have seen their search profile deleted instantly in the middle of an investigation by social media anti-bot AI rule. Once an investigator’s search profile is deleted, it is not possible to access social media data. Furthermore, the AI can now penalize an IP address, preventing new profiles from being used on those machines for up to 30 days.

2021 has been an important year for social media, and the venues have responded with their best technical answers. As social media data becomes more and more useful for insurance investigations, it is more critical than ever to check the work product delivered by social media software and teams. For those using automated social media solutions, it’s time to double check that the solutions still locate profiles accurately. It is also important to ensure whole profile is being delivered instead of just a handful of posts that could be captured before the AI closed the down the bot.

Social media data is stronger than ever as a method to document lifestyle, injury recovery, and ongoing behavior. If your investigation team is current with the new technical environment, the data will be robust. We all just have to work a little smarter to get it. 

By: Marci De Vries-Todtz, CEO, Fraud Sniffr Inc.