Sep. 26, 2023, Repost from Forbes, by Michael Nicosia, COO and co-founder at Salt Security. -- In this summer’s movie Oppenheimer, the potential to end World War II motivated the team at Los Alamos when building the atomic bomb. Beyond simply having such a weapon in the arsenal, the need to have it first—as a critical means of defense—was equally significant.
Getting there first matters to beat your opponent. That truth also applies to the world of cybersecurity—especially when it comes to artificial intelligence (AI). Cybercriminals are already leveraging AI, which puts organizations under increased pressure to build a strategy to combat these new threats. AI has the potential to not only accelerate the pace and number of attacks, but also create entirely new risks and AI-based attacks.
Generative AI tools like ChatGPT seem like a harmless novelty. You input a request—perhaps rewriting an article or building an itinerary for an upcoming trip—and within seconds, the algorithm generates a response. However, in much the same way, these tools can also be used for nefarious purposes, including writing phishing emails, performing data extortion and other scams.
In the U.K., AI has been classified as a “chronic risk” to national security in the 2023 National Risk Register. Meanwhile, many organizations, including Samsung, Amazon, JPMorgan Chase, and Apple, have already banned or restricted the usage of ChatGPT on business devices.
Restrictions may work to a degree, but they are never foolproof.
Adopting AI To Gain An Early Advantage
Security leaders understand the gravity of AI. In our recent global survey of CISOs/CSOs, respondents cited AI adoption as the top global trend most impacting their role—above macroeconomic uncertainty, geo/political climate and layoffs.
Security experts know that AI can easily pinpoint vulnerabilities for cyberattackers, making their ability to infiltrate or misuse systems that much simpler—and maybe worse, allowing them to learn from and repeat successful breaches.
If you are not equipped as well as your adversary, you operate at a disadvantage. The increasing use of AI in cyberattacks makes it essential that organizations adopt AI-powered solutions to “catch” and stop AI-driven attacks with solutions that tap AI to deliver increased situational context and adaptive intelligence.
Using AI As A Potential Deterrent
Given the stakes at hand, organizations must take measures now to adopt AI-driven security solutions. The growing AI cybersecurity market reflects this need. According to Acumen Research and Consulting, in 2021, the market for AI in global cybersecurity was $14.9 billion. By 2030, it is projected to reach a whopping $133.8 billion.
An attacker looking at two potential victims will focus on the one with the weakest defenses—where they know they can be successful without any discovery or danger. Attackers also know that AI can be used against them—through identification and legal reprisals. In the case of international spying, retribution could include political sanctions or counterattacks.
Only AI can scale to provide fast analysis of the ever-increasing volumes of data being produced in today’s digital-first economy. AI has the ability to find the needle in the haystack—something that isn’t feasible for humans. AI can quickly see if there have been changes within the environment and surface them.
The need to apply AI in cyber defense extends to another oft-discussed topic today: application programming interfaces (APIs). It’s estimated that 83% of web traffic is API traffic. Digital transformation runs on APIs, which enable the flow of information across today’s modern applications.
But APIs can also provide access to volumes of valuable financial and personal data that bad actors can harness for ransom, identity theft and fraud. To protect these vital conduits of data to online applications and services, organizations must be first—ahead of the malicious actors—to use AI to protect them. With the rapidly growing API ecosystem, only AI-driven API security has the ability to quickly monitor and analyze vast amounts of data to determine what API behaviors are good and what API behaviors are bad.
When choosing an AI-driven security product, companies need to consider the solution’s maturity. AI algorithms provide more value through time and exposure derived from a crowdsourced experience. Immature AI algorithms frequently generate significant false positives and false negatives. A proven AI solution will lower false positives and provide more extensive insights about potential security threats.
The importance of AI for cybersecurity defense cannot be underestimated. Moreover, harnessing AI cyber now—being among the early adopters—might just help organizations avoid being the victim of the next data breach.
Learn more about Salt Security.