Security needs to make its voice heard in AI implementations

Artificial intelligence (AI) is advancing at a remarkable rate — a phenomenon to behold, but also daunting for practitioners left to deal with the impacts of a technology that is outpacing the needed guardrails for responsible implementation. Whether it is the development of AI-specific frameworks, the need for more professionals who assess AI through a lens of ethics and digital trust, or the need for AI regulations, we’re not far enough along — and yet, the business imperative to leverage AI becomes more apparent with each passing day.

No company wants to be left behind, but proceeding without conducting due diligence opens the organization to massive risk. This holds true for multiple use cases: AI implementations to innovate product offerings as well as the use of AI in a cybersecurity context. To close some of these gaps in the short-term, there are important conversations to be had with executive leaders, and security professionals are well-positioned to start the dialogue. Among the important questions to raise:

• What are the goals of leveraging AI in our security program and are our implementations furthering those goals?
• Which AI-based security tools are we using and how do we know they are trustworthy?
• What is our organizational policy on usage of generative AI?
• How are ethical concerns about AI implementation surfaced and addressed?

While these and other questions might be thorny to resolve, they cannot be ignored because there is no question AI is a promising force on the security landscape. AI’s assistance in areas such as network intrusion detection, preventing phishing attacks and offensive cybersecurity capabilities can provide enterprises with major benefits. AI can significantly enhance cybersecurity by detecting and analyzing patterns in vast amounts of data, helping to identify potential threats and vulnerabilities faster and more accurately than traditional methods. Artificial intelligence and machine learning can also help organizations improve their data classification, which is especially useful in an era with increasing crossover between security, privacy, legal and compliance. Additionally, AI can be used to develop advanced defense mechanisms that adapt to evolving threats, making it harder for attackers to breach systems and respond to AI-powered threats. And let’s not forget the speed factor — AI can significantly enhance cybersecurity by detecting and analyzing patterns in vast amounts of data, helping to identify potential threats and vulnerabilities faster and more accurately than it takes people — no small thing, considering that time is of the essence in the security realm.

All of these advanced capabilities are urgently needed — partially because the power of artificial intelligence is already being tapped by the adversary, who will not wait around for organizations to work out the kinks before experimenting with how AI can serve their interests. Just as AI can be used to defend against cyber threats, it can also be leveraged to launch sophisticated attacks. AI-powered malware and bots can autonomously probe and exploit vulnerabilities, making cyberattacks more scalable and challenging to counter. Ceding AI usage to cybercriminals would give bad actors a major advantage. In many instances, the only way to effectively counterbalance the sophistication of AI-powered attacks is for security practitioners to harness AI in their defense methods.

As powerful as AI is, though, it lacks extraordinarily valuable human ingredients such as creativity, empathy and the ability to fully contextualize the meaning behind the data it is processing. That is why it is critically important that AI implementations complement the human ingenuity of security leaders and practitioners. In the rush to win the AI arms race between security professionals and cybercriminals, it is imperative that organizations don’t overlook the critical thinking skills that should determine when and how AI is deployed, and how it might impact the company’s reputation.

As AI expert Raef Meeuwisse explains, “The challenges for cybersecurity professionals will change from a primarily technical battlefield to an increasingly ethical and managerial one, forcing a redefinition of roles and responsibilities. It will no longer be sufficient to be technologically adept; future cybersecurity professionals will need to grapple with the philosophical and ethical dimensions of AI.”

To leverage the advantages of AI in cybersecurity while effectively addressing its potential risks, a comprehensive approach centered around proper governance, transparency and ethical principles is imperative. Implementing clear guidelines and frameworks for AI usage ensures responsible and accountable practices, fostering trust and confidence in AI-driven security solutions. Promoting collaborative efforts among researchers, policymakers and industry experts is indispensable in maintaining a cutting-edge defense against ever-evolving cyber threats. This collective approach enables the sharing of knowledge, best practices and innovative solutions, fortifying our collective ability to safeguard digital ecosystems and preserve the trust of individuals and organizations alike.