COVID-19 Creates Need For A Web-Based AI News Accuracy Tool

Datamation content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More .

I was going through my news this morning, and an alert caught my eye. According to the folks at Atlas VPN who have been monitoring the web, there are some terrifying statistics. As is sadly always the case when a disaster happens, bad people move to take advantage of the related fear.

Statistics

For instance, back in February, they identified 50 suspicious sites tied to COVID-19; currently, they are tracking 35K suspicious sites with sustained growth of 1,000 new suspicious sites per day (a whopping 1,900% increase).

Amazon has removed over half a million suspicious product listings, Shopify has closed 4,500 shady sites, Interpol is tracking $2M in related deceptions, and they are investigating a single case where someone lost $100K. In the UK alone, their people have lost $1.2M to date, and Interpol has so far blocked 18 bank accounts and frozen $703K in fraudulent transactions.

And we are still early on with this pandemic. Microsoft just significantly improved the security features of their Edge Browser against things like ID and Password theft, tracking, and privacy. Still, I’m not aware of anyone yet that has released a feature that can protect you from false information or commerce sites with fake goods. Even the majors like eBay and Amazon are struggling to keep up.

This problem would be an ideal use for a Deep Learning AI solution we don’t yet have, but that could be critical to protecting our employees, customers, and families from this overwhelming effort to take advantage of this Pandemic illegally.

Let’s talk about that this week.

The Problem

The issue is that people are panicked about the COVID-19 virus and the constant reports on death rates. No one wants to die, but they also need to get out to do things like grocery shop, and many don’t live alone so there is the chance they could catch the virus even if they don’t leave the house.

This kind of fear tends to increase confirmation bias; you make the decision first then do the evaluation ignoring warnings and negatives. The result is that scams that otherwise wouldn’t be very successful are getting results, and no amount of notification will be sufficient because the victims’ bias will overcome the warning in the notification.

So the potential victim needs a tool that can notify them accurately about the avoidable mistake they are about to make – and does so in a way that will get the user’s attention and change their behavior.

Why Deep Learning AI?

I’ve been looking into Deep Learning AIs ever since Deep Instinct, and HP hooked up with what arguably is the most robust PC security solution currently in the market. The reason is, like other AI AV solutions, it focuses on the behavior of a user or hostile code to determine intent and mitigate rather than having to be explicitly programmed to block a user or known malware.

Criminal behavior has patterns; once someone figures out how to execute a scam, others learn to emulate the approach, and while there will be variances between scammers, at the core, the goal and process to achieve it are very consistent. The scammer plays on the fact they know more than the victim and, by learning about the victim and saying things the victim wants to hear, they scam the mark out of their hard-earned cash and then move to the next victim until shut down.

An AI that could both look at the behavior of these scam sites and monitor the functioning of the potential mark to send out alerts or even automatically block the related transaction until the validity of the website can be confirmed. And an AI, which will learn about the victim and about other victims that have been scammed, can use that knowledge to present a warning that will have a higher probability of stopping the risky user behavior.

And these victims aren’t just end-users, they are also corporate and government buyers who are being scammed, which is why the actual damage numbers seem unusually high for single events.

Wrapping Up: The Critical Need For An AI Focused On Buyer Protection

This issue isn’t a spot problem because people are scammed every day by sellers who seem legitimate but aren’t. The need for an AI that can protect buyers just gets more pronounced during times of disaster because the disaster’s pressure tends to make potential victims less prudent.

As a result, the market needs an AI that is focused on keeping both individual and primarily corporate buyers from being scammed and, before we are done with COVID-19, I bet one of the big AI companies like IBM, Microsoft, or Google will bring one to market. It can’t come soon enough.