The Pros and Cons of Using AI in Cybersecurity

10.09.23 10:02 AM
Artificial intelligence (AI) is having an undeniable impact on the world, including the cybersecurity landscape. Cybercrime is a serious threat to businesses of all sizes across practically every industry, and many are turning to AI tools to help reduce risk, identify malicious activity, and stop breach attempts.

But do the benefits of using AI in cybersecurity outweigh the drawbacks, or are the downsides negligible when compared to the bigger picture? If you’re asking questions like those, here’s what you need to know about the pros and cons of AI in cybersecurity

The Pros of Using AI in Cybersecurity

Faster Identification of Anomalous Behavior

AI is particularly good at pattern recognition, so it’s often quick to identify anomalous activity within an environment. As a result, AI tools can potentially recognizing actions or happenings that indicate a hack is in progress or that one may occur soon far faster than a person would. That allows companies to take those insights and respond more quickly, making it possible to stop a malicious actor before a hack occurs or minimize the damage if an attack is in progress.

Immediately Respond to Events

Along with detecting anomalous behavior, AI can potentially respond to an incident without direct interaction from a person. For example, if it spots a log-on attempt from a device in an unexpected location – such as another country where the company doesn’t have a presence – it could log off the suspicious user, disable a current password, and notify administrators of the event for further action. 

Spot Zero-Day Attacks and New Malware

Identifying zero-day attacks is particularly challenging, as the vulnerability isn’t always widely known. However, AI can use a variety of tools and techniques – including machine learning (ML) – that may allow it to spot activity that could indicate a zero-day attack is occurring or possible. Similarly, an AI may notice activity that indicates the presence of previously unseen malware. That could allow swifter action to occur, ensuring that new malware is addressed quickly to minimize damage or ongoing risk.

Receive Response Recommendations

Even if a company is uncomfortable with AI cybersecurity tools taking action on their own, AI can still prove useful. These solutions can produce recommendations regarding next steps if they detect suspicious behavior. That can then guide the efforts of IT team members tasked with addressing the concerning activity, potentially speeding up the time it takes to resolve the problem. 

The Cons of Using AI in Cybersecurity

Resource Intensive 

While AI can reduce the workloads of cybersecurity professionals, it often requires a substantial amount of computing resources to accomplish its tasks. For companies with limited resources, that can prove problematic. Using the AI tool could make resources unavailable for other daily activities, diminishing productivity. Alternatively, it could make investing in more resources a must, and that could come with a significant price tag.

False Positives

While AI solutions are highly capable, they’re not perfect. Some AI systems may incorrectly label activity as suspicious and then alert cybersecurity team members to an issue that isn’t actually a problem. If the AI is also programmed to take action on its own in response to anomalous behavior, then using the AI could create more work, not less.

Vulnerable Data Sets 

AI technologies are typically only as strong as the data sets used to train them. While most AI developers take precautions when incorporating data into these training libraries, they aren’t entirely impervious to interference or inaccuracy. That means there’s always a chance that an AI could generate inaccurate or biased results. Additionally, it’s potentially at risk of manipulation, both at the beginning and as ML causes it to adjust its behavior. 

Privacy Issues

Many AI-based systems are continuously learning, and they use current user input to grow and adapt. While that can help keep AI cybersecurity tools current, it can also lead to privacy concerns. The maker of the AI solution may be gathering data from the company’s using its AI systems. While their intent may not be malicious, that does create a potential data privacy issue.

Derek Roush