5 Industry Leaders Share Their Bold Predictions for 2026
The technology landscape is evolving faster than ever, and 2026 promises to be a pivotal year for AI infrastructure and cybersecurity. We gathered predictions from five industry experts who are shaping the future of enterprise technology. Here’s what they see coming.
Moshe Tanach, CEO & Co-Founder of NeuReality: Data Centers Become the New Computer
The fundamental architecture of computing is changing. According to Tanach, developers will stop thinking about individual servers and start programming against pools of distributed resources. AI applications will be built as collections of microservices running across many server nodes, forming one logical system where locality is no longer a primary concern.
Ethernet will emerge as the dominant fabric unifying the entire data center into a single programmable computer, spanning both scale-up and scale-out networking. But the bigger shift? Organizations will move from adopting AI to optimizing it-focusing on real ROI, gross margins, and driving AI closer to a zero marginal cost model. This economic shift will disrupt existing business models and create space for new leaders who know how to run AI profitably at scale.
Denny LeCompte, CEO of Portnox: The CISO Job Gets a Reality Check
The Chief Information Security Officer role is undergoing a fundamental transformation. For at least a decade, companies hired CISOs to be paranoid and prevent every possible breach. But boards are now asking a different question: “How do you make breaches survivable?”
LeCompte predicts three connected trends: CISOs will stop promising perfection and focus on survivability; passwordless authentication will cross 50% enterprise adoption as the accepted standard for secure access; and cyber insurance will force a major reckoning, demanding proof of phishing-resistant authentication and documented micro-segmentation. Insurers already offer 15-25% premium reductions for proven micro-segmentation, and some are denying claims when MFA can’t be verified at the time of breach.
Mark Sokol, CMO of CyberFOX: The Death of Zero Trust (For Small Organizations)
Zero trust may be the ultimate goal, but for organizations with 2,000 employees or less, it’s becoming a barrier to actual security. Sokol predicts the rise of “practical resilience”-small, simple cybersecurity solutions that can be implemented quickly without massive resources or budgets.
The reality? Hackers aren’t looking for the hardest challenge; they’re looking for easy targets. Small businesses and government entities that chase perfect zero trust platforms often fail due to complexity and lack of resources, leaving them vulnerable today while pursuing tomorrow’s nirvana. Instead, implementing simple safeguards now makes you more secure immediately, causing attackers to move on to easier victims.
Richard Phillips, VP of Technology and Infrastructure at BackBox: AI as Trusted Advisor, Not Autonomous Agent
AI in security has arrived, but Phillips warns against the instinct to automate everything everywhere as fast as possible. The next era of IT security isn’t about replacing engineers with algorithms-it’s about building a trusted advisory layer.
His framework for AI strategy centers on three principles: transparency (knowing exactly how AI reached its conclusion), human-centricity (AI normalizes messy data so skilled humans can decide), and safety over novelty (avoiding friction and downtime in production). The leaders who win in 2026 won’t have the most autonomous tools; they’ll have the most explainable security programs. You can’t truly secure what you don’t fully understand.
Jeremy Brown, CTO at GitGuardian: The Explosion of Non-Human Identities
AI isn’t creating new security problems-it’s exposing the ones that were already there. GitGuardian found 24 million secrets leaked on public GitHub last year, up 25%. But private repositories are 9 times more likely to contain secrets than public ones because the safer people feel, the more careless they get.
The game-changer? Anyone can build apps with AI now-ops teams, sales, marketing-and they’re shipping straight to production without security training. Repositories using AI assistants have a 40% higher leak rate. Attackers have figured this out: why hack in when you can just log in with leaked credentials?
But the bigger challenge is non-human identities. Every AI agent, automation, and service needs credentials. Each person will soon have multiple AI agents acting on their behalf, and there’s no way to scope their access-agents must inherit whatever permissions the human has. We’ve spent years building systems to manage human access, but we’re way behind on machine identity governance. The challenge for 2026? Solving this without slowing down AI adoption.
##
