Machine-Speed Cyber Defense
Muratcan Koylan@muratcan
The idea
I believe AI is compressing the window between a vulnerability going public and it being exploited, from weeks down to hours, which breaks the old model of scan-ticket-patch. This basket leans into companies built for automated detection, response, and identity verification for both people and AI agents, anchored by larger cloud security platforms, with a small volatile sleeve for post-quantum encryption. I skip legacy scanning vendors on purpose, keep some cash on the side, and plan to hold for years unless the automation thesis itself stops playing out.
What it owns
- Attacks operating at machine speed require automated detection and response systems that act without human analyst intervention to be effective. · SentinelOneIt detects and responds to attacks autonomously using AI, with its Singularity ASOC platform automating security operations center functions that previously required human analysts., NetskopeIt provides cloud-native secure access service edge architecture designed for the AI era rather than retrofitted onto legacy hardware, competing directly with older approaches., Varonis SystemsIt automatically identifies sensitive data and restricts access to it, becoming more critical as AI agents gain the ability to read through company files unsupervised., A10 NetworksIt protects applications and network traffic, and is moving into specialized defense for AI applications while they run in production environments.22%
- As AI agents proliferate across enterprises, systems that verify and control identities for both humans and machines expand their serviceable market significantly. · OktaIt handles login and permission controls that AI agents need to operate safely within enterprise systems, making agent identity management a natural extension of its core business., SailPointIt governs access controls for human and non-human identities including service accounts and AI agents, which increasingly outnumber human users in enterprise environments., RubrikIt creates tamper-proof backups and enables rapid recovery of identity systems, serving as a last line of defense when automated attacks penetrate other security layers., Mitek SystemsIt verifies identities by analyzing government documents and detecting deepfakes, protecting against AI-generated fraudulent credentials that defeat human review.22%
- Large cloud-native security platforms already embed AI into their products, providing stable exposure to the machine-speed defense theme while smaller positions mature. · CrowdStrike HoldingsIt uses AI to automatically detect and stop attacks across endpoints and cloud systems, and recently expanded partnership with NVIDIA to secure AI agents specifically., ZscalerIt inspects every connection through its own cloud network rather than relying on perimeter security, an architecture essential when attacks move at machine speed., CloudflareIt distributes security fixes across a global network within hours and has rolled out post-quantum encryption at scale, addressing both near-term exploits and long-term cryptographic threats., Palo Alto NetworksIt consolidates security tools through acquisitions to close capability gaps, including defenses for AI agents, betting that it can adapt faster than legacy competitors.22%
- When exploits can be weaponized within hours of disclosure, companies that continuously find and fix vulnerabilities in code and supply chains gain structural advantage. · JFrogIt scans software packages in build pipelines to catch malicious or vulnerable open-source components before deployment, preventing supply chain attacks at the source., Tenable HoldingsIt is shifting from periodic vulnerability scans to continuous, AI-driven exposure management that constantly identifies attack surface rather than relying on scheduled assessments., GitLabIt finds and fixes security flaws in code before deployment using AI to suggest remediation, reducing the window between vulnerability creation and exploitation., DatadogIt monitors cloud applications around the clock, enabling real-time attack detection as threats unfold rather than discovery during post-incident analysis.17%
- Quantum computers will eventually break current encryption standards, making early vendors of quantum-resistant alternatives small, high-conviction bets on that transition. · SEALSQ CorpIt manufactures chips with quantum-resistant encryption embedded at the hardware level, representing an early commercial offering in post-quantum cryptography., BTQ TechnologiesIt develops post-quantum cryptography technology and is moving from development to commercial customer adoption as enterprises prepare for quantum-era encryption replacement., Arqit QuantumIt provides quantum-resistant key exchange technology for organizations preparing defenses against future computing threats.7%
- Cash10%