Weekly Wisdom: Software Reset and Structural AI Winners

Software’s Drawdown: Sentiment Shock, Not Structural Collapse

The software segment of the market has taken its largest non-recessionary drawdown ever, falling more than 30% from its September peak. More than $2 trillion in market capitalization has been erased, and the group’s weight in the S&P 500 has declined from 12.0% to 8.4%.¹ Concerns surrounding new large language model capabilities pushed sentiment to deeply depressed levels. The fear is that AI agents could coordinate enterprise workflows, automate complex tasks, and ultimately displace traditional seat-based software models, reducing pricing power and strategic relevance.

We believe the selloff reflects a sharp shift in investor sentiment rather than a sudden deterioration in fundamentals. AI innovation is moving quickly, which naturally increases uncertainty and warrants a higher risk premium. The market’s approach has been clear: sell first, ask questions later. The key debate is not whether AI changes software, but which layers are disrupted versus reinforced. Seat-based subscription models may face pressure as outcome-based and usage-based pricing models emerge. However, deeper layers of the stack are not easily bypassed. Enterprise software remains embedded within global corporations, protected by multi-year contracts, high switching costs, regulatory complexity, and proprietary datasets. The most probable outcome is additive integration rather than wholesale replacement. Current pricing appears to be discounting a full structural collapse, which may be overly aggressive.

Cybersecurity: AI’s Most Durable Beneficiary

While parts of software face valuation resets, cybersecurity is emerging as a structural beneficiary of AI adoption. Rather than reducing the need for security tools, AI is expanding the attack surface, increasing workflow complexity, and accelerating vendor consolidation. Enterprises are already responding. The expansion of chatbots, AI agents, and LLM-based applications is driving increased cybersecurity spending, with expectations for above-trend growth across monitoring software, cloud workload security, and data-loss prevention. According to recent surveys, 92% of CIOs and CISOs anticipate that Agentic AI will lead to increased cybersecurity spending in 2026, with 38% expecting a significant increase and 54% expecting a modest increase.² The cybersecurity landscape is becoming more complex, introducing new risks such as data exposure and expanded attack surfaces alongside persistent threats like phishing and malware, with nearly 47% of organizations citing adversarial advances powered by generative AI as their primary cybersecurity threat, enabling more sophisticated and scalable attacks.³

One development to correct the increasing cybersecurity risk is utilizing quantum computing technology. Asymmetric cryptography, the system for encrypting sensitive data today, is projected to be unsafe by 2030, prompting cybersecurity leaders to begin multiyear migration plans toward post-quantum cryptography.⁴ Organizations are being advised to start by building comprehensive cryptographic inventories using network security appliances, lifecycle management tools, or application security posture management platforms. One such organization is the Department of Defense, mandating in a November 2025 memo that all legacy cryptology components are to be phased out by December 2030, allocating $7.1 billion for the migration effort. Hybrid certificates may provide temporary compatibility, but classical signatures will not be supported indefinitely. This represents a structural, multi-year investment cycle rather than a short-term upgrade.

At the same time, generative AI is exposing structural weaknesses in traditional data loss prevention architectures. Encrypted traffic exposure, intent blindness, and shadow AI usage require adaptive, risk-based frameworks integrating Secure Service Edge (SSE), Data Security Posture Management (DSPM), and AI Security Platforms (AISP). These systems deliver context-aware protection, centralized visibility, and dynamic enforcement across increasingly complex AI workflows. Identity and access controls are becoming foundational as enterprises deploy autonomous agents at scale. As AI adoption expands, security becomes a strategic enabler rather than a discretionary spend.

Early Monetization of AI Cybersecurity

Recent results from Palo Alto Networks (PANW) provide a practical example of these dynamics. Despite near-term optics relating to guidance, the underlying indicators reflect consolidation strength as management notes that monetization remains early. Implementation of cybersecurity is ramping up as the company reported phantomizations in 2025 totaling 1,450, up 35% year-over-year. Additionally, net retention among platformed customers stands at 119% with low churn, underscoring vendor consolidation trends as enterprises streamline security stacks in response to AI-driven risk. XSIAM, PANW’s AI-driven security solution, surpassed $0.5 billion in ARR with over 600 customers. Secure Access Service Edge (SASE) exceeded $1.5 billion in ARR, growing approximately 40% year-over-year, while secure browser adoption reached over 9 million licenses.⁵ Management’s view is that large language models enhance security platforms rather than replace them. In a world where AI increases speed and complexity, scaled and integrated cybersecurity platforms appear to be gaining strategic importance rather than losing it.

The Creator Economy: Technology Is Expanding Opportunity

There has been growing concern that AI will erode job opportunities and weaken the consumer. Yet the data suggests the opposite dynamic may be unfolding. Goldman Sachs forecasts that the global creator economy will reach nearly $500 billion by 2027, roughly doubling since 2023. Social commerce now represents 19% of global e-tail sales, with two-thirds occurring in China. Influencer advertising spending reached approximately $37 billion in 2025, growing 26% compared to 6% growth for the broader media industry.⁶

New forms of monetization are emerging alongside digital platforms. MrBeast, the most followed creator on YouTube, with his self-branded chocolate company Feastables generated approximately $250 million in sales last year, producing more than $20 million in profit.⁷ The future opportunity in markets is not connected to manpower, but audience ownership and brand equity. A compelling example is the influencer, Logan Paul. Five years ago, he purchased a collectible Pokémon card for approximately $5 million and then created content around the category, hosting unboxings, reviews, and integrating the asset directly into his personal brand. He ultimately wore the card in a custom diamond-encrusted necklace at WrestleMania, transforming it from a niche collectible into a global cultural moment. The card was later sold for roughly $16 million, setting a record for the most expensive trading card ever sold at auction.⁸ By leveraging his audience and brand, Logan Paul highlighted how influencers can meaningfully impact demand, pricing, and the future of marketing itself.

More broadly, AI is lowering the barrier to entry in content creation and enabling more individuals to harness brand equity and audience attention at scale. The broader takeaway is that AI is not eliminating economic activity; it is reshaping it. Advertising dollars are reallocating, entrepreneurship is expanding, and digital platforms are enabling new income streams that traditional employment metrics often fail to capture.

Stephanie Link’s TV Schedule

---

[1] Bloomberg, as of February 18, 2026
[2] Mizuho: CISO/CIO Security Survey, as of February 17, 2026
[3] Google: 2025 Global Cybersecurity Outlook, as of May 1, 2025
[4] Garner Invest: Cybersecurity Trends, as of February 10, 2026
[5] Palo Alto Networks: Q4 Earnings report, as of February 17, 2026
[6] Bloomberg Intelligence: as of December 3, 2024
[7] Bloomberg, as of February 17, 2026
[8] Bloomberg, as of February 17, 2026