Digital fraudsters have He never succeeded. The Americans lost last year $ 16.6 billion for online crimesOf almost 200,000 people reporting fraud, such as phishing and fbi falsification. Over $ 470 million was stolen in fraud, which began with a text message last year, According to the Federal Trade Committee. As the largest mobile manufacturer of operating systems in the world, Google tries to do something by building tools to warn consumers from potential fraud.
Before starting Google Android 16 next week, the company said on Tuesday that it is so Expansion his Recently launched AI flag function In the case of the Google Message application, known as detecting fraud, in order to ensure notifications of potentially wicked messages, such as possible cryptographic fraud, impersonating financial, gift card and fraud prizes, technical support fraud and many others. In combination with other AI security functions for Google messages – all of which work locally on user devices and do not provide data with data or messages – and now detects about 2 billion suspicious messages per month.
“The fraud is really painful,” says Dave Kleidermacher, vice president of engineering at the Android Security and Privacy Department. “There is really a very huge amount – the most important epidemic and scourge for humanity – financial fraud that is all over the world.”
Fraudsters operate all over the world, but Chinese fraud groups are especially for millions of dishonest messages, demanding such things as payments of “fees” or information about alleged supply of postal services. When people click the links and enter their data, including information about payments, fraudsters steal their data. In some cases, frauds are designed as a kind of Smash-Anda-Grab, in which the attackers quickly deceive users to resign from the crumbs of information, such as a pair of logins or a credit card number. These frauds are usually more formal and potentially easier to detect. A more convoluted challenge is the detection of highly involved investment or romance fraud – often called pigs – which build and evolve for months of messages, while fraudsters build relationships with goals before they save them to transfer their savings to life, and even debt to send more money.
“Get to fraud – not only click the link,” says Kleidermacher. “With artificial intelligence, you can actually watch and watch these more sophisticated conversations, and then detect their fraud.”
Thanks to the kindness of Google
Thanks to the kindness of Google
On the screenshot of the fraud detection function provided by Google, the encrypted RSC chat shows the typical news of the fraud, in which the fee for the payment for EZ is excellent. The message adds that “legal capacity” to keep can be withdrawn if the payment is not made. The message contains a link that directs someone towards a malicious payment. The fraud detection overlay at the bottom of the screen says that “suspicious activity” has been detected in a message and offers a way of reporting and blocking the sender, along with an option that allows people to indicate that this is not a fraud.
Meanwhile, in recent months, the finish has begun, which is the owner of WhatsApp, Messenger and Instagram Introduce pop -up warnings When people are asked to make payments in chat messages. Elsewhere, Cybersecurity Firma F has created a beta tool to facilitate people identify whether the message and sender are Probably fraudsters and block messages. Placing a layer of friction, which moves people away from messages that they do not know, or responding to messages with a request for details, can reduce the chances that fraudsters will be successful.
Kleidermacher Google says that the company sees “a really positive impact” of using machine learning systems to detect potential frauds in real time. As the collateral matures, he notes that the basic system can ultimately spread outside the Google Messages application on the communication platforms of other companies.
For now, some of these extensions start with their own Google products. The company also said on Tuesday that in the early stages of testing methods of enabling fraud detection in the case of telephone connections, but the possibility was not widely implemented.