Microsoft uses AI to stop hackers from scamming you on your phone

Developed by Microsoft,
Azure Operator Call Protection is a new service that scans for suspicious phone interactions and alerts users of fraudulent calls.

14 is the number of unwanted calls a user receives per month. This is also the result of a study conducted by voice security platform Hiya which analyzed 98 billion phone calls around the world. A global scourge that recently alerted the United States Federal Communications Commission to such an extent that it took the bull by the horns to counter this spam, by declaring fraudulent calls using voices generated by artificial intelligence illegal.

It is in this context and on the occasion of the MWC in Barcelona that Microsoft presented its new service, Azure Operator Call Protection. Imagine. You receive a call that you choose to answer. Immediately, Microsoft’s AI analyzes the conversation in real time and identifies its fraudulent potential. In the event of spam or cold calling detected, you are notified directly by the operator. All you have to do is end the conversation. It’s as simple as an AI.

A service that makes you master of the game

AI analyzes calls in real time and looks for specific indicators of fraud, such as the use of insistent language or being encouraged to divulge sensitive information over the phone.

It then interrupts the call to inform the user as soon as it detects possible fraud. This is when the user has the choice to end the call immediately or get more details about the reason for the report. In other words, the system does not limit itself to disconnecting the call in the event of suspected fraud; the final decision is up to the user. By providing all available information, the tool then gives the user the freedom to make their own decision.

You remain in control from the beginning to the end of the call. The service is otherwise “opt-in”, requiring user consent, and no saving or use of call data will be made to power Microsoft’s AI. “ Once the call ends, and once the customer has chosen to listen to the recommendation or make additional requests, there is nothing more said Shawn Hakl, vice president of 5G strategy for Microsoft’s Azure for Operators program.

Azure Operator Call Protection acts as a sentinel, alerting users and helping protect them against phishing attempts and other forms of fraud. And if hackers do not lack imagination to impersonate banks or large brands, the artificial intelligence model will evolve to adapt to the emergence of new forms of threats. In other words, this AI turns its opponent’s weapon against him.

A vast program to fight against scams

Microsoft has partnered with BT Group to test and refine Azure Operator Call Protection technology. This collaboration is essential to evaluate the effectiveness of the service in a real environment and to identify potential areas for improvement. The marketing date has not yet been specified, but ongoing tests will determine the feasibility and timetable for large-scale deployment.

Microsoft’s initiative is part of a broader fight against phone scams, which are a growing problem globally. Various actors are tackling this problem using different approaches. AT&T, for example, implemented a legitimate call identification system to help users distinguish between fraudulent calls. Last November, the White House also hosted a virtual hackathon to encourage the development of artificial intelligence technologies that can detect spam and unsolicited calls.

The scope of Azure Operator Call Protection may expand beyond voice calls in the future. “ We evaluate voice calls first [….] But there is obvious interest in the text as well said Shawn Hakl.

So learning about AI is only just beginning.

Source: CNET, Hiya