Join the event trusted by corporate leaders for almost two decades. VB Transforma connects people building AI Real Enterprise. Learn more
AI Enterprise infrastructure expenses are expected $ 309 billion By 2032, the winners will not be determined by who has the best models; It will come down to the one who controls the infrastructure layer, which makes AI work on a enormous scale.
Security suppliers make the most aggressive movements. Palo Alto NetworksIN Crowdstrike AND Cisco Each report security revenues based on AI-naic artificial intelligence increases by 70 to 80% year on year, while established sales of infrastructure have dropped. The pattern is clear: security becomes a control plane for AI Enterprise.
“The complexity of AI loads blows up the existing infrastructure to a breakthrough point”, Ali Ghodysi, general director of Databicks, notices in the blog post. “Enterprises need basically new approaches to AI management on a large scale.”
The evidence is approaching. According to Idc73% of enterprises give non -ex -infrastructure as the main barrier in the AI party. Meanwhile, opponents of AI weapons faster than enterprises can implement defense. Infrastructure wars began.
Agenticops appears as a modern battlefield
Agenticops is not a vision of one supplier. It is a recognition in the entire industry that established IT operations cannot manage AI agents operating at machine speeds with human powers. Cisco started the category at Cisco Live 2025But AI Microsoft orchestrationIN Google operations and startups like Weights and prejudices They are already racing to have it. The battle lines are drawn.
Technical requirements are brutal. Companies implementing 50,000 AI agents need infrastructure that supports access to data between domains, real -time management and cooperation with many teams. Established tools crack at 5000 agents. Mathematics does not work.
“For the first time, security becomes accelerating to adoption, not obstacles,” said Venturebeat Jeetu Patel, president and Cisco, said Venturebeat. The change is fundamental: security teams now allow you to implement AI instead of blocking it.
Three pillars define corporate class agencies: unified access to data in all domains, cooperation environments in which Netops and Secops teams work together, and specially built models regulating agent activities. Forrester Research It confirms the visibility of many domains as critical. Sellers who master all three components will dominate. But most have difficulty with effective delivery of one.
Death of peripheral safety
Established firewalls cannot protect AI loads. The evidence is overwhelming. Palo Alto’s Prisma Cloud Provides 2 billion safety events daily during the performance. Fortinet’s Security Fabric It connects over 500 integration points because the peripheral defense failed. Infinity control of the point It works on the principles of zero trust, assuming a violation in every layer.
The extended Berkeley (EBPF) filter has changed the game. This Linux testicle technology allows you to enforce safety without 40% of established approaches. The acquisition of $ 2.8 billion isovalental cisco confirmed this approach. SnapshotThe Open Source Isovalent project now protects the production loads at Netflix, Adobe and Capital One. 15,000 GitHub stars reflect the adoption of enterprises, not the interest of developers.
Craig Connors, vice president of Cisco and Cto Safety, formulated a change in a recent interview with Venturebeat: “Security policy currently has the use of each layer, from burdening work to silicon.” Implication is clear. Safety becomes an integral part of the infrastructure, not an overlay.
Acceleration of equipment seals the transformation. Protected silicon safety works in nanosecond delay. Mathematics is brutal: the security defined by the software adds 50-200 milliseconds. Hardware safety adds 50 to 200 nanoseconds. This is a million-times improvement. Suppliers cannot compete without silicon possibilities.
72-hour exploit window
Opponents of the gap in resolution within 72 hours. Patch Enterprises in 45 days. This gap generates 84% of successful violations. Each security seller races to close it.
Falcon Crowdstrike prevents Exploits blocks before the existence of patches. Qualys VMDR Provides real -time sensitivity management. Patch Tanium He promises an automated reaction. Cisco’s Live Protect uses testicular levels in a few minutes.
Economics are undeniable. Studies Ponmon Institute show that every hour of delayed patching costs USD 84,000 at risk of violation. Automated platforms provide a return on investment (ROI) in 4.7 months. Ciso cannot ignore mathematics.
“Time is everything in cyber security,” emphasizes Shlomo Kramer, CEO Cato networks. “Automation is not only performance; it is about surviving attacks that human teams cannot react quickly enough.”
Observable wars intensify
$ 28 billion SLLUNK Acquisition signals a greater truth: observation determines who wins the battle of AI infrastructure. DataDAND 18 trillion event processes per day. New relic It monitors 10 billion transactions per minute. Dynatatatrace It follows 2.5 million applications in the cloud.
The rates are existential. Enterprises implementing artificial intelligence without observation fly blindfolds. “You can’t secure what you can’t see,” says ETAY Maor, senior director of the security strategy at Cato Networks. “Obserability is not optional, it is the basis of safe digital transformation.”
The user interface represents the next border. Instead of AI navigation desktops, it creates real -time interfaces based on an correct problem. Servicenow, Splunk and Raising Players Noticeable They bet that energetic interfaces replace inert navigation desktops within 24 months.
Market consolidation accelerates
Giants of infrastructure gather their armies by acquiring. Cisco paid $ 28 billion for a splant. Palo Alto has acquired the safety of cider, digging security and a voucher for a total of $ 1.2 billion. Crowdstrike bought reposifify, Humio and prejudiced. The takeover of VMware worth $ 69 billion in Broadcom transforms the entire landscape.
The platform speed now determines survival. Unified architecture reduces development time from years to months. What took 18 months on implementation now begins within 8 weeks. Engineering teams vote with their legs, joining the companies that send at the speed of entrepreneurship at the speed.
The AI infrastructure market is expected to consolidate from over 200 suppliers to less than 20 platforms within 36 months. Gartner predicts By 2027 60% of current suppliers. The message is brutal: control a full stack or becomes irrelevant.
Lower line
Agenticops is the most significant architectural change since the cloud processing appears. Enterprises that build AI infrastructure, assuming a continuous compromise, infinite identities and machine speed attacks, will develop. People attaching peripheral defense and human speed reaction will join the blockbuster and Kodak at the digital cemetery.
Suppliers solving this challenge – be it Cisco, Palo Alto, Microsoft or emerging players – will control the next decade of corporate technology. The race is turned on. The clock is ticking. The winners are already appearing.
