As 2019 gets started, people are naturally looking for ways to improve current processes as technologies improve.

For example, some researchers are investigating ways to depend on artificial intelligence (AI) to crack down on cybercriminals including hackers and scammers. Here are four ways the technology might help make gains this year:

1. To Protect Networks Against Entirely New Threats

The antivirus and malware software that most people know works by recognizing known threats. That works well to a point, but it means there’s always a chance hackers could come up with an attack type that an installed tool doesn’t know yet.

A company called Darktrace wants to do things differently by relying on AI to pick up on unusual system traffic and make smart judgments about what constitutes threats.

It can do that even if the attack is affecting the system for the first time. In one recent case, Darktrace prevented the network of a Singaporean finance company from getting infiltrated by malware embedded in an email that an employee opened. If the AI software were absent, the hackers would have been able to gain substantial access to the system. But, Darktrace noticed the strange activity and stopped it.

Such proactiveness could prove crucial in stopping hackers from doing their planned damage — havoc that can often be costly. Research shows that a cyberattack on a small business costs between $84,000 and $148,000, and the same report showed that approximately 60 percent of small businesses cease operations within six months of a cyberattack or breach.

Darktrace and other companies that deploy AI solutions could signal a new trend in cybercrime prevention where the tools go beyond screening for the threats seen before.

2. To Shield People From Spam Calls

Robocalls are becoming increasingly prevalent and annoying. Some people get several per day and have stopped answering their phones in frustration. Although not all unwanted calls are spam, data from YouMail shows that of the 4.4 billion robocalls Americans received in September 2018, 40 percent were spam.

But, Google’s fighting back with a feature on its Pixel phones. It checks calls on behalf of a phone owner. A person can activate it by tapping a “screen call” button. Then, the caller hears a message that the phone owner is using a call screening option from Google and asks them to state their name and the reason for the call.

The Google Assistant — which uses AI — then transcribes what the caller says and sends the transcription back to the recipient, letting the person decide whether to take the call.

Although robocallers target physical phones, they often fit the description of an online scammer, using special software to contact people quickly through single computer interfaces.

Technology has improved enough to make these calls so efficiently that the scammers don’t mind if many individuals in a given market don’t fall for the tricks. The volume they can make in a short period makes it worth their while regardless.

3. To Spot Forged Emails

Some scammers impersonate genuine companies and insert branded graphics to trick recipients into revealing details or otherwise falling for a scam.

Another attack, called a business email compromise (BEC), occurs when a hacker impersonates a known executive. A new algorithm aims to cut down on BECs by analyzing more than 7,000 writing characteristics and comparing them against an AI model trained with a person’s real style.

If the algorithm notices discrepancies, it sends a warning to the implied sender as well as a company’s IT department. Most BEC attacks reportedly come from an individual who impersonates the CEO or a company’s managing director. The fraudulent emails then often contain urgent requests, such as to perform wire transfers.

Some lottery scammers operate in similar ways. Tips for avoiding them warn people to beware of getting asked to provide their bank details to set up a wire transfer. Instead of transferring a person’s “winnings” to the account, scammers take all the money out.

4. To Detect Credit Card Fraud

Statistics about global cybercrime estimate that it costs the world’s economy approximately $600 billion. And, one of the most common but preventable kinds of cybercrime is credit card fraud, especially as people are increasingly eager to buy things online.

Companies have developed AI tools that spot unusual transactions with algorithms that understand the typical types of purchases for certain customers and compare them to other “clusters” of people making similar ones.

Like the AI product mentioned above that highlights suspicious network traffic, these fraud-detection tools learn what’s out of the ordinary and raise the alarm when needed.

One thing that makes this approach superior to the frameworks that flag anything falling outside a given set of parameters is that people get placed into new clusters based on the things they buy.

For example, if some of those transactions indicate they are taking a vacation soon, they’d get put in a group of other people planning vacations, and alerts would only arise if a person’s behavior deviates too much from others in the cluster.

Cybercriminals Will Likely Use AI, Too

These developments undoubtedly show promise for helping people avoid cybercrime, whether at the enterprise or individual level.

But, as AI gets smarter and more practical to use, it’s inevitable that those who orchestrate cybercrime will try it to achieve their means.

So, people depending on AI to stop attacks must ensure their algorithms stay as up to date as possible to stay ahead of the hackers and scammers.

For more market trends, visit ETF Trends.