By Kayla Matthews
Cybersecurity is one of the biggest ongoing global challenges. It’s not uncommon for companies to experience thousands or even millions of attempted cyberattacks each year. Adding connected infrastructure and IoT devices to the mix is a priority for many businesses, but these technologies broaden the threat surface substantially.
Thankfully, technology continues to deliver solutions for technology’s potential harm. AI and threat protection is, maybe not surprisingly, a potent combination in this area. Here are some of the ways artificial intelligence is improving how we protect ourselves against digital threats in 2020 and beyond.
1. AI Continues to Improve the Speed of Threat Detection
Predicting and responding to cyber threats in the “traditional” manner is time-intensive, costly and extremely difficult. The required tasks run the gamut of network topography, from detecting illegitimate connections and requests to identifying large packets of data or unusual programs that come online after an enterprise network connection is established.
Artificial intelligence is far better than human beings at determining which portions of web applications, mobile platforms, and database management tools are legitimate and which might be a malicious actor. AI shifts the company’s responsibilities and responses from the extremely difficult “predict and respond” model to the “detect and respond” model.
A string of high-profile acquisitions of AI threat protection companies proves that this technology has staying power. In 2020, financial institutions are using AI at a greater scale than ever to comb huge datasets of transactions and compare the results against past and real-time results: an essentially impossible task for human employees as banks and credit unions add customers and grow in size.
2. AI Can Spot Flaws in Software and Improve Bug-Hunting
Organizations need to focus on lots of moving parts if they want to keep themselves protected against digital threats. One of these moving parts is updating critical hardware and software as soon as patches become available. Artificial intelligence can streamline the asset management process by identifying which products have pending updates and applying them automatically at the right time.
Artificial intelligence is set to dramatically change how our software is written, too. AI can comb thousands of lines of new code much faster than programmers can to find and flag common or new vulnerabilities that a cybercriminal could exploit.
Presently, NIST’s library of known “Common Weakness Enumerations” contains more than 10,000 entries, each of which represents a potential security breach if it slips through the cracks of the QA process.
3. AI Helps Automate Cloud Security Management
The digital age presents many opportunities alongside its risks. Remote working opportunities and mobile computing using the cloud is one of the most useful.
In a business environment, however, it can be difficult to tell when a network or cloud system may have been corrupted by improper or out-of-date credentialing. In-house data breaches are some of the most common and damaging types, representing almost one-third of all cybersecurity incidents.
Remote work environments, where employees exchange messages and documents remotely, can be some of the most vulnerable to cybercriminals. But AI can improve how organizations “police” their geographically distributed computing systems and shared file repositories by identifying cases of improper credential usage and unauthorized data access. AI makes it easier to detect when unauthorized parties use credentials they shouldn’t have or access data they shouldn’t need for their work.
In some cases, AI can even be used on an ongoing basis to determine the legitimacy of a network user based on personal behavioral identifiers like typing style and speed.
4. AI Improves the Vetting Process for Website Visitors
Cybersecurity is a major struggle for eCommerce companies and others that rely on an extensive internet presence to reach customers and stay in touch with existing ones. Ensuring that website visitors and customers are who they say they are is an important part of keeping online shopping portals and digital services free from bad actors.
AI will help improve the user authentication process throughout 2020, especially compared to existing verification systems like Captcha. Captcha is good for weeding out automated attempts to connect to a service or network, but it doesn’t identify whether the owner of the credentials is the actual owner of the account.
AI can help by detecting legitimate, returning users on websites. For eCommerce sites and other digital properties, AI can be used to “read the signs” and verify the identity of users as they log into existing accounts, versus bad actors logging in using stolen credentials.
5. AI Will Make Future Networks Like 5G Safer
There will be around 21 billion internet-connected devices active in the world by 2025. Moreover, this Internet of Things will be supported by the next generation network technology known as 5G. 5G is available only in a select handful of locations, but it will take the capabilities of our personal and industrial IoT devices to the next level of speed and usefulness.
With AI, 5G will also deliver far improved safety over previous technologies. According to Huawei, adding AI to 5G infrastructure can improve the speed of threat detection by 87%. What this means is that AI is an ideal companion for keeping wireless industrial control systems and enterprise planning software running uninterrupted, the way it should be, without forced outages or downtime caused by security problems like ransomware or DDoS attacks.
Ransomware, in particular, is a very real threat in healthcare and industrial settings, where bad actors can lock employees out of diagnostic equipment or other assets until they pay a ransom. In time, AI could put these threats to bed for good by eliminating the database model of threat identification. Instead, AI uses machine learning to keep ahead of potential threats — even new threats the system hasn’t recognized yet and for which there is no database entry.
AI and the Next Generation of Threat Protection
AI-based threat protection is not a cybersecurity “silver bullet,” but it’s definitely a game-changer. There is no excuse not to use robust antivirus programs, firewalls and intrusion detection software. Each of these is still a vital component of any complete cybersecurity package. But we can expect AI to continue to improve the effectiveness and refine the features of these familiar digital security products.
AI is coming of age at an ideal time, too. In total, cybercrime will cost the world economy some $5.2 trillion of its value between 2019 and 2023. Recovering that value means building our company cultures on security-mindedness and using advanced threat protection technologies — including AI.