Cybersecurity shares advance as AI safety debate widens

Cybersecurity stocks jumped on Wall Street on Monday as warnings over the pace and control of increasingly autonomous artificial intelligence systems drove investors away from semiconductor shares and towards companies expected to benefit from heavier spending on digital defence.

CrowdStrike closed about 13.9% higher and Palo Alto Networks gained roughly 13.1%, placing both among the strongest performers in the S&P 500. Intraday gains approached 15% for CrowdStrike and 14% for Palo Alto Networks, while Nvidia, AMD and other AI-linked chipmakers fell as investors reassessed the consequences of calls for slower frontier-model development.

The breadth of the move extended beyond the two leaders. Zscaler, SentinelOne, Okta and Fortinet also climbed sharply, while the iShares Expanded Tech-Software Sector ETF rose more than 5%, highlighting the day’s unusually wide software outperformance overall.

The market rotation followed a sharpening debate among leading AI developers over whether capability gains are outrunning safeguards. Anthropic chief executive Dario Amodei has argued that the world should preserve the option to slow or temporarily pause frontier AI development if safety systems and institutions cannot keep pace. OpenAI has also said it would slow or stop development or deployment when safeguards are inadequate, after identifying cybersecurity capabilities and model misalignment as areas requiring stronger controls.

That shift has placed cybersecurity companies at the centre of a new investment thesis: more capable AI can strengthen defenders, but it can also automate reconnaissance, coding, credential theft and intrusion at a speed that compresses the time available to respond.

CrowdStrike’s 2026 Global Threat Report said operations by AI-enabled adversaries increased 89% year on year. It put average eCrime breakout time — the interval between initial compromise and lateral movement — at 29 minutes in 2025, with the fastest observed case taking only 27 seconds.

Software security data point to a parallel risk on the development side. Veracode’s 2026 GenAI Code Security Report found an average security pass rate of about 56% across the AI models it tested, meaning roughly 44% of tasks produced code containing a detectable vulnerability. The report said syntactic correctness had approached 100%, underscoring a widening gap between code that runs and code that is secure.

CrowdStrike is responding with SafeMind, a family of purpose-built cybersecurity models introduced this month and designed to operate inside its Falcon platform. The company says the system combines offensive and defensive models in a shared loop, using telemetry from its security platform rather than relying solely on general-purpose frontier models.

Its financial momentum has also strengthened. CrowdStrike reported second-quarter fiscal 2027 revenue of $1.47 billion, up 26% from a year earlier, while annual recurring revenue rose 25% to $5.84 billion. Net new ARR reached a record $333 million, prompting the company to raise its full-year growth outlook for that measure.

Palo Alto Networks has taken a more acquisition-led route to the same market. It has added businesses focused on AI gateways, agentic endpoints and security automation, including Portkey, Koi and Console, while integrating the larger CyberArk acquisition into a broader platform strategy spanning network, cloud, identity and security operations.

For its fiscal fourth quarter, Palo Alto Networks reported revenue of $3.41 billion, up 34% year on year, and remaining performance obligations of $21.2 billion, also up 34%. Chief executive Nikesh Arora said advances in AI were pushing cybersecurity higher on chief information officers’ priority lists and strengthening demand for real-time protection.

The two companies therefore enter the AI-security cycle with different operating models. CrowdStrike is emphasising internally developed, security-specific AI layered onto Falcon, while Palo Alto Networks is combining organic development with acquisitions intended to fill gaps across the emerging agentic security stack. Their latest results show strong growth at both companies rather than a clear winner, leaving execution, integration and customer adoption as the key differentiators.