Adversaries from cybercrime gangs to nation-state cyberattack squads are fine-tuning weaponized AI with the purpose of defeating new patches in 3 days or much less.
The faster the assault, the extra time to discover a sufferer’s community, exfiltrate information, set up ransomware or arrange reconnaissance that can final for months or years. Conventional, guide patching is now a legal responsibility, rendering enter organizations defenseless in opposition to weaponized AI assaults
"Risk actors are reverse engineering patches, and the pace at which they're doing it has been enhanced significantly by AI," Mike Riemer, SVP of Community Safety Group and Subject CISO at Ivanti advised VentureBeat in a latest interview. "They're in a position to reverse engineer a patch inside 72 hours. So if I launch a patch and a buyer doesn't patch inside 72 hours of that launch, they're open to take advantage of."
This isn't theoretical hypothesis. It's the exhausting actuality forcing distributors to rearchitect their safety infrastructure from the kernel up fully. Final week, Ivanti launched Join Safe (ICS) model 25.X, marking what Riemer calls "tangible proof" of the corporate's dedication to assembly this menace head-on.
At DEF CON 33 researchers from AmberWolf proved this menace actual, demonstrating full authentication bypasses in Zscaler, Netskope, and Test Level by exploiting vulnerabilities that existed for months, together with Zscaler's failure to validate SAML assertions (CVE-2025-54982), Netskope’s credential-free OrgKey entry, and Test Level’s hard-coded SFTP keys exposing tenant logs have been all flaws left open and exploitable greater than 16 months after preliminary disclosure.
Why Kernel Safety issues
The kernel is the central orchestrator of every part that occurs in a computing gadget, controlling reminiscence, processes, and {hardware}.
If an attacker compromises the kernel, they've seized whole management of a tool that may scale to compromising a whole community. Another safety layer or software, platform or safeguard is straight away bypassed with attackers take management of the kernel.
Practically all working programs depend on the idea of imposing rings of privilege. Functions run in person mode with restricted entry. The kernel operates in kernel mode with full management. When adversaries break that barrier, they’ve gained entry to what many safety researchers contemplate the holy grail of a programs and full networks’ vulnerabilities.
Ivanti's new launch straight addresses this actuality. Join Safe 25.X runs on an enterprise-grade Oracle Linux working system with sturdy Safety-Enhanced Linux (SELinux) enforcement that may restrict a menace actor's skills throughout the system. The answer contains Safe Boot safety, disk encryption, key administration, safe manufacturing facility reset, a contemporary safe net server, and Net Software Firewall (WAF), all designed to safe key elements of the system and considerably deter exterior threats.
"Up to now 12 months, we've considerably superior our Safe by Design technique, translating our dedication into actual motion by way of substantial investments and an expanded safety crew," Riemer defined. "This launch stands as tangible proof of our dedication. We listened to our prospects, invested in each expertise and expertise, and modernized the safety of Ivanti Join Safe to supply the resilience and peace of thoughts our prospects anticipate and deserve."
From OS rings to Deployment Rings: A extra full protection technique
Whereas working system rings outline privilege ranges, fashionable patch administration has adopted its personal ring technique to fight the 72-hour exploit window.
Ring deployment supplies a phased, automated patching technique that rolls out updates incrementally: a Check Ring for core IT validation, an Early Adopter Ring for compatibility testing, and a Manufacturing Ring for enterprise-wide rollout.
This strategy straight addresses the pace disaster. Ring deployment achieves 99% patch success inside 24 hours for as much as 100,000 PCs, in response to Gartner analysis. Ponemon Institute analysis reveals organizations take an alarming common of 43 days to detect cyberattacks even after a patch is launched.
Jesse Miller, SVP and director of IT at Southstar Financial institution, emphasised: "When judging how impactful one thing will be, you need to take every part from present occasions, your trade, your setting and extra into the equation." His crew makes use of ring deployment to scale back their assault floor as rapidly as attainable.
Attackers aggressively exploit legacy vulnerabilities with 76% of vulnerabilities leveraged by ransomware have been reported between 2010 and 2019. When kernel entry is at stake, each hour of delay multiplies the chance exponentially.
The Kernel Dilemma facilities on balancing safety versus stability
At CrowdStrike's FalCon convention, Chief Know-how Innovation Officer Alex Ionescu laid out the issue: "By now, it's clear that if you wish to shield in opposition to dangerous actors, it’s essential to function within the kernel. However to do this, the reliability of your machine is put in danger."
The trade is responding with elementary shifts:
-
mandates multi-year modifications for each Home windows safety vendor
-
for safer kernel instrumentation
-
Apple's Endpoint Safety Framework
permits user-mode operation
Authentication bypass occurs when kernels are compromised
AmberWolf researchers spent seven months analyzing ZTNA merchandise. Zscaler didn’t validate SAML assertions (CVE-2024-54982). Netskope's authentication could possibly be bypassed utilizing non-revocable OrgKey values. Test Level had hard-coded SFTP keys (CVE-2025-3831).
These vulnerabilities existed for months. Some distributors patched quietly with out CVEs. As of August 2025, 16 months after disclosure, many organizations nonetheless used exploitable configurations.
Classes realized from compressing 3 years of kernel safety into 18 months
When nation-state attackers exploited Ivanti Join Safe in January 2024, it validated Ivanti’s choice to quickly advance its kernel-level safety technique, compressing a three-year venture into simply 18 months. As Riemer defined, "We had already accomplished part one of many kernel-hardening venture earlier than the assault. That allowed us to rapidly pivot and speed up our roadmap.”
Key accomplishments included:
-
Migration to 64-bit Oracle Linux:
Ivanti changed an outdated 32-bit CentOS OS with Oracle Linux 9, considerably lowering recognized vulnerabilities tied to legacy open-source parts.
-
Customized SELinux enforcement:
Implementing strict SELinux insurance policies initially broke a big variety of product options, requiring cautious refactoring with out compromising safety parameters. The ensuing resolution now runs in everlasting enforcement mode, Riemer defined.
-
Course of de-privileging and safe boot with TPM:
Ivanti eradicated root privileges from essential processes and built-in TPM-based safe boot and RSA encryption, making certain steady integrity checks, aligning with AmberWolf’s analysis suggestions and findings.
There have been additionally a collection of unbiased penetration testing initiatives, and every confirmed zero profitable compromises, with menace actors usually abandoning makes an attempt inside three days.
Riemer defined to VentureBeat that international intelligence neighborhood prospects actively watched menace actors probe the hardened programs. "They tried previous TTPs, pivoted to net server exploits. They beautiful a lot gave up after about three days," Riemer mentioned.
The choice to go kernel-level wasn't a panic response. "We truly had plans in place in 2023 to handle this earlier than we ever received attacked," Riemer mentioned. The dialog that sealed the choice occurred in Washington, DC. "I sat down with the CIO of a federal company, and I requested him flat out: Is there going to be a necessity for the U.S. authorities to have an L3 VPN resolution on-prem sooner or later?" Riemer recalled. "His response was that there would all the time be a mission want for an L3 VPN on-prem kind resolution with a purpose to give encrypted communication entry to the warfighter."
The long run past kernel safety contains eBPF and Behavioral Monitoring
Gartner's Rising Tech Affect Radar: Cloud Safety report charges eBPF as having "excessive" mass with 1-3 years to early majority adoption. "Using eBPF permits for enhanced visibility and safety with out relying solely on kernel-level brokers," Gartner notes.
Nearly all of cybersecurity safety distributors are investing closely in eBPF. "At this time, virtually our complete buyer base runs Falcon sensor on prime of eBPF," Ionescu mentioned throughout his keynote at this 12 months’s Fal.Con. "We've been a part of that journey as eBPF basis members."
Palo Alto Networks has additionally emerged as a significant participant in eBPF-based safety, investing closely within the expertise for his or her Cortex XDR and Prisma Cloud platforms. This architectural shift permits Palo Alto Networks to supply deep visibility into system calls, community site visitors, and course of execution whereas sustaining system reliability.
The convergence of CrowdStrike, Palo Alto Networks, and different main distributors on eBPF expertise alerts a elementary transformation—offering the visibility safety groups want with out catastrophic failure dangers.
Defensive methods which are working
Patching is usually relegated to a kind of duties that will get procrastinated about as a result of so many safety groups are short-handed, dealing with continual time shortages. These are the circumstances that adversaries financial institution on after they select victims.
It’s a positive wager that if an organization just isn’t prioritizing cybersecurity, they are going to be months and even years again on their patching. That’s what adversaries search for. Patterns emerge from totally different industries of victims and so they share a standard trait of procrastinating about system upkeep normally and safety patterns particularly.
Primarily based on interviewing victims of breaches that began with patches generally years previous, VentureBeat has seen the next speedy steps they take to scale back the probabilo9ty of being hit once more:
Automate patching instantly. Month-to-month cycles are out of date. Tony Miller, Ivanti's VP of enterprise providers, confirmed ring deployment eliminates the reactive patching chaos that leaves organizations susceptible throughout the essential 72-hour window.
Audit kernel-level safety. Ask distributors about eBPF/ESF/WISP migration plans and timelines.
Layer defenses. That is desk stakes for any cybersecurity technique however essential to get proper. "Whether or not it was SELinux profiling, root privilege avoidance, an up to date net server, or the WAF—every layer stopped assaults," Riemer mentioned.
Demand transparency. "One other vendor had been attacked in November 2023. That data didn't come obtainable till August 2024," Riemer revealed. "Because of this Ivanti has been so public about transparency."
The underside line
Kernel-level transformation isn't optionally available. It's survival when AI weaponizes vulnerabilities in three days.
Ivanti Join Safe 25.X represents what's attainable when a vendor commits totally to kernel-level safety, not as a reactive measure, however as a elementary architectural precept. Gartner's strategic planning assumption is sobering: "By 2030, a minimum of 80% of enterprise Home windows endpoints will nonetheless depend on hybrid endpoint safety brokers, growing the assault floor and requiring rigorous validation."
Organizations should harden what they’ll now, automate instantly, and put together for architectural upheaval. As Gartner emphasizes, combining ring deployment with built-in compensating controls together with endpoint safety platforms, multifactor authentication, and community segmentation as a part of a broader zero-trust framework ensures safety groups can shrink publicity home windows.
