With the top of 2025 shortly approaching, it is time to sit up for 2026 and discover some key themes safety leaders ought to concentrate on to assist focus their efforts.
In reviewing my 2025 predictions, one which exceeded my expectations was Palo Alto Networks’ acquisition of CyberArk. In December 2024 I wrote, “For each zero belief and SASE, integrations with identification suppliers are crucial. Privileged entry administration has develop into a core use case for a lot of zero-trust community entry distributors. In the end, I may see a situation the place a community safety vendor appears on the identification market and decides there’s sufficient alternative to take an opportunity.” I am unsure I had a $25 billion acquisition in thoughts once I wrote that.
So, what’s in retailer for the upcoming 12 months? Unsurprisingly, AI has at the least a point out throughout all my predictions for 2026. Which is not to say it is the one factor on the agenda, nevertheless it has develop into so pervasive that it is going to be at the least a tangential consideration in most areas.
With out additional ado, listed here are my 5 community safety predictions for 2026.
Zero belief for AI turns into a precedence
Zero belief has shifted from a advertising and marketing buzzword to a safety crucial, bolstered by reference architectures and government orders from the federal authorities. Many organizations are implementing broad zero-trust architectures throughout their environments, however because of the complexity of those initiatives and the truth that zero belief is a journey and never a vacation spot, it stays a piece in progress for many.
But, IT innovation continues to march ahead. Earlier than most organizations have a deal with on zero belief typically, they’re going to have to contemplate tips on how to apply it to AI. This contains securing person entry to public AI purposes, AI mannequin entry to knowledge sources and, most significantly, AI agent entry to a wide range of sources.
Id is already a foundational pillar of zero belief, however many organizations are nonetheless engaged on incorporating risk- and contextual-based entry fashions to their environments. This shall be crucial, particularly as agentic architectures take maintain and develop into extra autonomous.
Dwelling-off-the-land assaults develop into pervasive, partially attributable to AI
LOTL assaults have been an ongoing pattern over the previous couple of years, with Bitdefender reporting that 84% of the 700,000 cyberincidents it analyzed used LOTL methods of some variety. As attackers proceed to make use of AI and shift to agent-based assaults, the prevalence of LOTL assaults will solely develop.
Detecting and mitigating these assaults will take a multilayered and coordinated effort. Community-based instruments, similar to microsegmentation and community detection and response, shall be crucial to detecting and stopping lateral motion, command and management visitors, and different anomalous habits that bypasses signature-based defenses.
Browser safety heats up, however to reinforce, not exchange current instruments
The browser safety market has garnered lots of consideration over the previous couple of years. The seller aspect has seen vital exercise, and 2026 will see safety groups make investments closely in these capabilities.
Quite than changing current instruments, nevertheless, most groups will take a look at browser safety as a complementary addition to the instruments already in place. A part of this shall be pushed by the kind of approaches obtainable. SASE distributors have launched browsers to increase safety to the endpoint and assist AI safety. Standalone safe browser extension distributors tout the flexibleness of continuous to make use of a regular browser with safer capabilities. Not all organizations wish to deploy a brand new, standalone browser to all their customers. However there is no such thing as a query that including larger visibility and management over exercise within the browser must be on each safety chief’s precedence record in 2026.
The AI safety market begins to skinny out
This shall be a continuation of a pattern we noticed in 2025. Most AI safety firms concentrate on a selected AI use case, for instance, securing worker use of public AI purposes, defending inside AI fashions and AI-enabled purposes, securing agentic AI architectures and utilizing agentic architectures within the SOC. These use circumstances are too near current markets for established distributors to face pat, and too distinct for a unified AI safety strategy most often. Add within the pattern of platformization and it stands to motive that enormous community and SaaS safety distributors, software and API safety distributors, identification distributors and SOC distributors will discover it most possible to accumulate the capabilities they should lengthen totally into AI.
SaaS safety consolidation accelerates
Equally, it has develop into more durable and more durable for SaaS safety distributors to outlive as standalone entities. Now we have already seen notable acquisitions by Examine Level, Fortinet and CrowdStrike on this space. That is additionally arguably the realm the place AI is the cleanest extension of an current market. Workers accessing public AI purposes is extraordinarily near the normal SaaS safety use case — safety groups want visibility and management over what purposes are being accessed, the information being shared with these purposes, distinct insurance policies for company situations and open situations, and extra. Monitoring for malicious immediate returns and monitoring content material generated by these purposes are extra particular capabilities, however on the entire, they’re related use circumstances. We’ll proceed to see many of those capabilities be subsumed by SASE platforms.
John Grady is a principal analyst at Omdia who covers community safety. Grady has greater than 15 years of IT vendor and analyst expertise.
Omdia is a division of Informa TechTarget. Its analysts have enterprise relationships with know-how distributors.
