As artificial intelligence reshapes how businesses operate, security leaders are increasingly vocal about the risks and structural shifts emerging alongside its benefits. CEOs from Cloudflare, Zscaler, and CrowdStrike highlight a common theme: AI is accelerating change faster than organizations, regulators, and security models can comfortably adapt
Cloudflare CEO: AI Is Rewriting the Internet’s Business Model
Cloudflare CEO Matthew Prince warns that AI is increasingly positioning itself between companies and their customers. Instead of visiting websites directly, users are turning to chatbots and AI summaries, weakening brand visibility and reducing traffic-driven revenue. Even research-driven businesses feel the impact as investors rely more on AI-generated insights than original sources.
Prince points out that large AI models are consuming vast amounts of online data, pushing companies to seek exclusive or proprietary datasets to stay competitive. In response, Cloudflare now enables publishers to block AI crawlers that do not pay for access, giving content creators more control over how their data is used.
A key concern, according to Prince, is the lack of clear rules around AI data collection. He argues that blending search dominance with AI data scraping creates an uneven playing field. Separating search access from AI crawling, especially in policy frameworks, could encourage fair competition and attract publishers seeking predictable rules and compensation.
While licensing agreements with news outlets and platforms like Reddit exist, disputes over scraping and copyright continue. Ambiguous standards, Prince warns, risk undermining incentives to produce high-quality content. His near-term recommendation is pragmatic: publishers should technically restrict unpaid AI access, and regulators should enforce transparency, consent, and compensation for valuable datasets in an AI-driven web.
Zscaler CEO: AI Agents Are Supercharging Cyberattacks
Zscaler CEO Jay Chaudhry believes the biggest AI risk today isn’t theoretical—it’s already unfolding. Speaking to CNBC, he noted that AI-powered cyberattacks are evolving far faster than most organizations can respond. Enterprises, he argues, are slow to adapt, and this lag is becoming a major vulnerability.
The focus is shifting away from traditional generative AI models toward agentic AI systems—autonomous agents capable of reasoning, decision-making, and executing multi-step tasks without constant human prompts. These systems are now being used by both attackers and defenders, escalating the cybersecurity arms race.
For attackers, AI agents dramatically lower the skill barrier and enable large-scale attacks. For defenders, they compress response times and expose weaknesses in fragmented security architectures. Supporting this view, Chaudhry cites data showing that 76% of organizations struggle to keep up with AI-driven attack complexity, while nearly half of security leaders rank AI-powered threats as their top ransomware risk.
His warning is clear: AI agents are entering a dangerous phase quietly, and many organizations may not realize the severity until after significant damage has been done.
CrowdStrike CEO: AI Is Pushing Security Back to the Endpoint
CrowdStrike CEO George Kurtz sees AI adoption driving renewed demand for endpoint security. While endpoint detection and response (EDR) has always been central to CrowdStrike’s business, the rise of generative AI tools and AI-powered browsers has introduced a new wave of endpoint risk.
Employees are increasingly installing AI applications directly on their machines—tools like ChatGPT and Claude—boosting productivity but also expanding the attack surface. According to Kurtz, AI browsers further amplify this risk by introducing new vulnerabilities alongside new capabilities.
As AI workloads shift closer to users, the endpoint has become the focal point of both human and non-human interaction with AI systems. This shift has led to accelerated demand for CrowdStrike’s Falcon platform, including large-scale deployments such as a government agency replacing legacy antivirus software across tens of thousands of endpoints.
Kurtz’s message is straightforward: in an AI-driven world, endpoint security is no longer optional—it’s foundational.
Across these perspectives, a clear pattern emerges. AI is not just a productivity tool; it is fundamentally reshaping the internet, threat landscape, and security architecture. Whether it’s protecting content, defending against autonomous attacks, or securing endpoints, organizations that fail to adapt quickly risk being outpaced by both innovation and adversaries.