Multi-layered AI protection needed to deal with AI-driven threats

“AI adds a vital layer of automation in threat response and mitigation, helping teams to stay resilient without adding complexity,” shares Rohit Aradhya, VP Engineering & Managing Director, Barracuda India.

As AI changes the threat landscape, businesses in the Asia Pacific region are now finding it more difficult to manage their cybersecurity. While there continues to be significant investment in cybersecurity, the biggest challenge is being able to manage the many cybersecurity tools as well as ensuring employees practice cybersecurity hygiene.

According to Rohit Aradhya, VP Engineering & Managing Director, Barracuda India, cybercriminals today are using AI to scale, automate and improve their attacks, making threats such as phishing increasingly personalised, faster and harder to detect. In an interview with CRN Asia, Aradhya shared that businesses with limited resources in particular are struggling to keep pace with this evolution.

“Many have fragmented security environments, with too many security tools that don’t integrate, reducing visibility and adding management complexity. Most still have too many manual security processes, reducing their ability to respond to the speed of AI-powered attacks. It can be easy for organizations to feel overwhelmed and underprepared for AI-powered threats. It is therefore important to remember that security service providers and the cybersecurity solutions they depend on are also increasingly AI-powered and ready to counter these threats,” Aradhya said.

Interestingly, with increased AI usage in organizations, especially agentic AI applications, Aradhya believes that attackers are using agentic AI to progress their attacks and in how they target business agentic AI deployments.

“Autonomous AI agents are capable of running an attack campaign from start to finish with minimal human involvement. They could set objectives, gather intelligence, craft convincing lures, test multiple intrusions paths and learn from their environment, making real‑time decisions as an attack unfolds. Rather than following fixed scripts, they have the capability to observe how defences respond and adjust their tactics, timing and targets accordingly,” he explained.

Aradhya pointed out that when combined with specialized criminal platforms, such systems could analyse vast volumes of data to automatically identify exploitable weaknesses and generate malware that rewrites its own code, blends more easily into normal activity and evades both signature‑based and behavioural detection.

Agentic AI challenge

“Traditional, static security controls are far less effective in a world of dynamic, AI-powered threats. Defences will need to evolve quickly, with a stronger focus on behaviour‑driven detection that can identify abnormal activity as it happens, rather than relying on known indicators of compromise,” Aradhya explained.

At the same time, he acknowledges that agentic AI will reshape how organizations operate and defend themselves. In cybersecurity, Aradhya pointed out that agentic AI will take responsibility for tasks such as applying software patches, watching for signs of network instability and organising incident‑response actions.

“Many businesses are realising that today’s threats are faster, more targeted and sophisticated and their security can’t keep up. Our research shows that among Indian organizations, 71% say they are juggling too many security tools, and more than half, 57% say these tools don’t integrate with each other. A worrying 80% say this makes it harder to respond to and mitigate security threats,” he said.

Given these figures, Aradhya highlighted the need for advanced, integrated protection that reduces operational strain and response times is universal. This includes ensuring companies have security that can keep up with the speed of modern attacks while reducing complexity for resource‑constrained teams.

“Barracuda is helping them meet this challenge by delivering deep, multi-layered AI‑powered protection. Through the BarracudaONE platform, we bring together a broad set of integrated security capabilities in a single, centralised experience, giving organisations stronger defence. By combining human expertise with intelligent automation, Barracuda enables faster response, better visibility and more effective protection for businesses that need to stay secure in an increasingly automated threat landscape, “he said.

AI vs AI era in cybersecurity

For Aradhya, as AI is increasingly central to both attack and defence, more cyberattacks will leverage AI and the only way to counter them is with the same speed and sophistication.

“Organizations must prepare for threats that evolve faster, and incidents that progress faster than human teams can track. AI adds a vital layer of automation in threat response and mitigation, helping teams to stay resilient without adding complexity. Security operations will shift toward continuous, automated analysis. AI will handle more of the investigation and response workload, but for now human oversight will remain critical to guide decisions and manage risk,” Aradhya said.

In this sense, Aradhya believes the immediate future will involve AI tools interacting with one another, digital agents will operate quietly in the background, but humans will set direction and maintain control.

He also believes that the use of AI within organizations will also need to be secured. The use of unauthorized shadow AI, data leakage and misuse of sensitive information, will make strong data and AI agent security posture management essential.

Given the increasing complexity of threats, Ayadhya advices organizations that are unsure of how to respond to AI‑powered threats to double down on the security basics and not be afraid to bring in expert help to support. This includes leveraging managed security service providers that work closely with security vendors to provide advanced extended detection and response.

“AI presents a rare opportunity to rebalance cybersecurity’s long‑standing asymmetry, where attackers need to succeed only once and defenders must succeed every time. AI can automate defence, accelerate response and provide proactive protection so teams can focus on outcomes rather than complexity. This matters especially for organisations without dedicated security staff, who need solutions that work for and with them. The tools protecting your organization must evolve faster than the threats,” he concluded.