As AI reshapes enterprise technology strategies, it is also redrawing the boundaries of cybersecurity. What was once a battle of tools and controls is fast evolving into a high-speed, AI-powered arms race between attackers and defenders.
In conversation with iTNews Asia, Robert Pizzari, Group Vice President, Asia, Splunk explains how the convergence of AI and exploding machine data is today creating a new landscape that is redefining security operations, exposing gaps in governance, and forcing organisations to re-examine their readiness in the new threat landscape.
The use of AI has seen an unprecedented explosion of operational and machine data, a trend that shows no signs of slowing down in 2026. At the same time, tools like generative AI chat interfaces have led to a perception that intelligence can be easily layered on top of this data.
However, that assumption is increasingly being challenged. “Organisations that have a belief that AI is a magic pill might be somewhat disappointed until they work out that it is about the quality of data,” Pizzari said.
He added that the long-standing principle of “garbage in, garbage out” is proving especially relevant in AI-driven environments. Without clean, structured, and contextualised data, even the most advanced AI systems struggle to deliver meaningful outcomes.
AI has industrialised cybercrime
The rise of generative AI is not just transforming enterprise operations; it is also empowering attackers. According to a recent Splunk CISO report, about 95 percent of CISOs are calling out the increased sophistication of threats as one of the new challenges since the explosion of generative AI.
Pizzari emphasised that at the core of this transformation is the ability of attackers to weaponise AI and make them more convincing and harder to detect. He added that AI is now enabling the “industrialisation” of cybercrime, driven by three key factors including human-like, emotionally persuasive content, automation at machine speed and scale and multi-layered attack strategies.
In many cases, AI-generated phishing acts as the entry point into broader campaigns involving malware deployment and data exfiltration. “We only have to make one mistake… and that can lead to downstream compromise,” he said.
According to Pizzari, beyond external threats, internal risks are also mounting. particularly around unsanctioned AI usage.
“There’s another theme appearing… shadow AI, where employees may not be using approved tools,” he explained. “Once data is uploaded to these models, it’s very difficult to retrieve or delete, it’s essentially on public record.”
Without strict governance frameworks, organisations risk exposing sensitive data, often without visibility or control. This reinforces the need for robust data management practices alongside AI adoption.
Despite advances in automation, Pizzari stressed the importance of keeping humans in the loop. Human expertise continues to play a critical role in validating insights, identifying anomalies, and making judgment calls. AI systems, while powerful, are still prone to hallucinations, bias, and errors, especially when trained on imperfect data, he added.
Organisations must shift from security to improving resilience
As the threat landscape evolves, traditional security metrics focused on prevention and control are no longer sufficient. “We need to start by measuring digital resilience and not just security controls,” Pizzari said.
He emphasised that CIOs and CISOs must evaluate how quickly their teams can detect and respond to attacks, how effectively systems can recover from disruptions, and how resilient operations remain even under sustained threat.
For enterprise leaders to increase their resilience, he advised a path forward that balances both technology and people.
First, strengthen your governance and guardrails for AI deployment. Second, invest in unified data visibility and AI-assisted detection. Third, it is important to develop your talent alongside technology.
- Robert Pizzari, Group Vice President, Asia, Splunk
Balance innovation with the need for governance and security
The emergence of AI has not just intensified cyber threats, it has also fundamentally changed their nature. What we are witnessing is not a temporary spike, but the beginning of a sustained cyber arms race, Pizzari explained.
For organisations, he said the challenge is twofold: keeping pace with increasingly sophisticated attackers; while ensuring their own systems remain secure, governed, and resilient.
“The winners in this new era will not be those who adopt AI the fastest, but those who deploy it most responsibly - balancing speed with control, and innovation with resilience,” Pizzari said.
He added that the evolution of the Security Operations Centre (SOC) will also be central to enterprise resilience, evolving beyond traditional monitoring into a data-driven, AI-augmented environment.
Pizzari explained that SOCs must focus on improving data quality and leveraging AI to reduce noise, accelerate detection, and enhance operational outcomes.
“The future SOC combines integrated workflows, automation, and AI-driven assistance with a firm “human-in-the-loop” approach, ensuring critical decisions remain guided by expertise while maintaining resilience, continuous monitoring, and strong guardrails against risks like data leakage, bias, and hallucinations,” he said.
