Outpacing the Adversary — Securing AI Innovation Before It’s Too Late

This was a sponsored session.

In an era where threats evolve with unprecedented speed and sophistication, IT and security leaders must rise to the challenge. As attackers harness AI, CISOs are pivotal in safeguarding their organisation without stifling innovation. 

Recently, CISOs in our DACH Community gathered for a Town Hall discussion on managing the rapidly evolving threat landscape when it comes to attacks using AI. Haider Pasha, Chief Security Officer, EMEA & LATAM at Palo Alto Networks, moderated the discussion. 

Ali Baccouche, Regional Information Security & Data Privacy Officer, EMEA at Texas Instruments; Daniel Maier-Johnson, CISO at KUEHNE+NAGEL; Judith Wunschik, Global Chief Cybersecurity Officer at Siemens Energy; and Rob Lyle, Principal Solution Architect, Security Specialist at Amazon Web Services served as panelists.
 

Highlights from the Discussion

The panel of security leaders began by acknowledging both the hype and the real opportunities presented by AI, particularly in the context of cybersecurity. The panelists agreed that their roles are evolving to help broker the benefits and risks of AI for their organizations, emphasizing that AI is not new but has reached a tipping point in its capabilities and impact. 

As one security leader put it, “AI is a technology for us to enable and improve decision-making as a business.” Another panelist agreed, saying “AI is part of a business-driven digital process.”

• Use cases for AI. Each panelist described how their organizations define and use AI. One executive highlighted AI’s role in enabling business decisions and operational flexibility, such as predictive maintenance. Other security leaders shared that they are leveraging AI to optimize supply chains and manufacturing. Another CISO is showing the value of AI in accelerating proof-of-concept work and extracting insights from data. In fact, one panelist explained that AI at their organisation is “a powerful way to get value from data.” The discussion underscored that AI is deeply embedded in digital transformation, with applications ranging from HR chatbots to manufacturing optimization. 
• Measuring value. The panelists also highlighted practical examples where AI has delivered measurable benefits. Multiple leaders described how AI-powered chatbots and automation have drastically reduced the burden of internal service tickets, freeing up HR and IT staff for higher-value work. Other executives emphasized AI’s ability to streamline processes, eliminate repetitive tasks, and improve customer satisfaction. The panelists cautioned that AI should not be adopted simply because it’s a buzzword—its impact on the bottom line and workforce efficiency must be clear. As one executive said, “We should expect a positive impact on the bottom line.”
• Challenges with AI. The conversation then turned to challenges, particularly the tension between security and agility, and the shortage of skilled personnel. The panelists agreed that as adversaries weaponise AI to launch more tailored and rapid attacks, defenders must also leverage AI to keep pace, especially in filtering out false positives and focusing on high-value security tasks. They noted that while AI can help address the skills gap by automating triage and analysis, a skilled workforce remains essential, particularly as responsibilities for AI governance and security increasingly fall to CISOs and security teams.
   

Key Takeaways for Implementing AI in Security Operations

• Strong AI governance and foundational security practices are essential. The panel stressed the importance of transparency, governance, and foundational security practices in AI adoption. Several executives highlighted that while AI accelerates detection and response, classic monitoring and a strong security culture remain vital. Proactive threat modeling using AI was seen as a key strategy for anticipating unknown risks, with panelists advocating for responsible ownership, risk assessment, and continuous improvement. The consensus was that while AI is transforming cybersecurity and business operations, it must be implemented thoughtfully, with robust guardrails and a clear understanding of its limitations and risks.
• Prioritize people and change management. The panelists emphasized the importance of focusing on the people impacted by AI adoption, not just the technology. Fostering relationships with stakeholders and considering the human element in change management are vital. AI can be leveraged for training, awareness, and operational efficiency, but successful implementation depends on clear communication, collaboration, and support for employees as roles and processes evolve.
• Start with clear problems and ethical foundations. When implementing AI, organizations should begin by identifying the specific problems they want to solve, rather than being swayed by vendor hype or generic AI solutions. Ethics and cultural context must be considered from the outset—models should be tuned to align with organizational values, compliance needs, and local regulations. Transparency about how models are trained and used is essential, and organizations should be proactive in building ethical guardrails, recognizing that ethical standards may differ across countries and industries.

These takeaways highlight that successful AI adoption in security operations requires a holistic approach—balancing governance, people, and ethical considerations with a clear focus on solving real business problems. As one panelist put it, CISOs are looking to “fight AI with AI.”

CISOs can continue the conversation on securing AI innovation at an upcoming gathering with their security peers. Community members can sign in to the app to find and register for events, or if you are new to Gartner C-level Communities, apply to join your regional CISO community to regularly connect with peers on critical priorities.
 

Special thanks to Palo Alto Networks and Amazon Web Services.

By CISOs, For CISOs®

Join the conversation with peers in your local CISO community.