New Content & Platform Update
New AI Knowledge Pass Program: Practical Introduction to AI: Claude Desktop
Release date: October 30th, 2026
What are we doing? The Program is nine hands-on labs. Each takes about 25 minutes, so participants can finish in up to 3 hours and 45 minutes. Topics include:
- Working with AI: when to use it
- Prompting with structure and examples
- Reusable assistants
- Bias in your prompts
- DLP in practice
- Shadow AI
- AI agents
- AI content: copyright, labeling, and disclosure
- Demonstrate your skills
Who is this for? Anyone in your organization who will use Claude Desktop. No prior AI experience is needed.
Availability This content is available to all Immersive customers!
New Lab Collection: Practical Introduction to AI – Claude Desktop
Release date: October 30th, 2026
What is this release?
Knowing what AI is isn't the same as using it safely, and the habits that matter most depend on the tool in front of you. Where saved instructions live, who can see them, what the assistant remembers, and which actions it can take without asking all work differently in each tool. As more organizations deploy Claude, learners need to practice these habits in the app they'll use every day.
The regulatory pressure hasn't changed. Since August 2026, national regulators can enforce Article 4 of the EU AI Act, which requires organizations that use AI to take measures to support their staff's AI literacy. Article 50's labeling rules for some AI-generated content have applied since August 2026 too. This collection gives organizations that use Claude practical, hands-on evidence of those measures. It also builds the habits that prevent the most common mistakes: oversharing data, using unapproved tools, and trusting AI output without checking it.
Who is this content for
This collection is built for the general workforce in organizations that use Claude Desktop, especially people who have completed Introduction to Using AI (either as part of the Knowledge Pass or as standalone content). Organizations that don't use Claude can choose the original Practical Introduction to AI collection, which covers the same skills.
New CTI Cyber Ranges Exercise - Iris C2Agent (LLM-Operated Implant)
Release date: October 30th, 2026
We’re releasing a new Cyber Range Exercise (CRE) to expand our coverage of emerging artificial intelligence (AI) threats: Iris C2Agent: Hunting an LLM-Operated Implant. This exercise directly supplements our recently released lab on the Iris implant, giving your teams a collaborative, realistic scenario to practice investigating LLM-driven intrusions.
Iris – first identified by Palo Alto Networks’ Unit 42 in September 2026 – is an advanced Windows implant where operators issue high-level goals in plain language instead of manual commands. The implant hands interactive host control over to a Large Language Model (LLM), which autonomously determines commands to execute, interprets terminal outputs, and decides subsequent actions. Recent variants can also hijack locally installed AI tools and abuse legitimate commercial AI services, blending malicious activity with normal outbound traffic.
While our foundational lab introduces individual practitioners to the relevant indicators of compromise (IOCs), this CRE allows security leaders to evaluate how their entire team handles a full-scope incident. Starting from a detection alert triggered by a suspicious domain name, participants must hunt through host and network logs, trace the adversary’s actions, extract the malware sample left behind, and conduct basic static analysis to understand the threat.
Who are we doing it for
Security analysts, incident responders, and threat hunters who defend Windows environments, engineer detection logic, and investigate multi-stage attacks.
Defenders operating in AI-enabled organizations who must learn how legitimate workplace AI tools and permitted external AI traffic reshape their external and internal attack surfaces.
Security managers and team leads looking to benchmark and improve their blue team’s readiness against autonomous, AI-driven malware.
Why are we doing it now
Iris represents an active, rapidly evolving shift toward autonomous, AI-operated malware. Because Iris can hijack approved enterprise AI tools, defenders face the critical challenge of differentiating between benign AI interactions and malicious command execution – especially since most corporate networks already allow AI web traffic by default.
Link to the exercise: here
New program release: 3x PHP2Shell Readiness Tracks
Release date: October 23rd, 2026
- PHP2Shell Readiness: Defensive Track
- PHP2Shell Readiness: Offensive Track
- PHP2Shell Readiness: Executive Decision-Making Track
Why are we doing it now
This release provides a ready-to-pitch framework, allowing security leaders to drive complete organizational alignment during a major crisis. Instead of training in isolation, your teams get a unified response playbook: hands-on engineers build the muscle memory to hunt indicators, patch runtimes, and prove exploitation paths, while C-suite leaders practice high-stakes trade-offs under real pressure.
Customer Availability
- Offensive & Defensive Tracks: Live today and available to all platform customers.
- Executive Decision-Making Track: Live today for all customers with Crisis Sim access (due to the program concluding with a crisis sim exercise scenario).
MITRE ATLAS Heat Map Update: Meet the AI Program Builder
Release date: October 22nd, 2026
Turn your AI security coverage gaps directly into action with our new Create program button. Located at the top right of the heat map, this feature builds a targeted learning plan for your team in seconds. You can generate programs automatically from your current filter settings targeting low and medium coverage gaps across your entire organization or focusing on specific teams, sub-teams, or coverage levels. Before publishing, refine your plan by instructing the builder to adjust difficulty (such as beginner or intermediate), set the duration and estimated hours, or filter by audience and specific techniques.
Who are we doing it for
Security leaders, team managers, and administrators on the Immersive One platform who need to eliminate AI security skill gaps fast.
Why are we doing it now
Identifying AI security gaps isn't enough! You need to close them immediately to reduce organizational risk. Stop wasting hours manually building custom training plans. By linking heat map insights directly to an automated builder, you can instantly convert gap analysis into tailored, actionable learning plans that keep your workforce ahead of emerging AI threats.
New Crisis Simulation Content: PHP2Shell Incident
Release date: October 22nd, 2026
We're launching a new crisis simulation called PHP2Shell Incident. Orchid Media designs and operates enterprise websites. It hosts 180 customer sites and delivers another 420 from customer-owned servers through its gateway. At 08:00 UTC an emergency bridge call opens on two critical PHP vulnerabilities that nobody has published yet. A leaked post has claimed remote code execution, PHP has confirmed a security release for 14:00 UTC, and the affected versions are still unknown. Orchid's own PHP inventory is incomplete, and the runtime data for the 420 customer-owned sites is only what owners have reported.
You work two seats as the evidence develops. The Web Platform Security Lead can change PHP and restrict routes across the hosted estate. The Incident Commander approves anything that takes customer services offline. That split is the point of the exercise. Orchid's emergency contract lets it suspend a customer's public gateway route without asking first, but only the owner can patch the server behind it. Play runs from the pre-advisory scramble through the patch release, overnight recovery, and a next-morning reassessment of what was missed.
It's based on the PHP security release of October 21, 2026, which fixed CVE-2026-109618 and CVE-2026-109619.
Who are we doing it for?
• Web platform, hosting, and infrastructure teams: people who own a runtime estate and have to decide what comes off the internet while the advisory is still pending.
• Incident commanders and IR leadership: the seat that signs off a precautionary customer outage and owns what customers get told about it.
• Vulnerability and asset management owners: the exercise turns on inventory quality, and it rewards teams who can say what they run and who owns it.
• Service providers and anyone with customer-owned estate: the central problem is holding risk you can see but can't fix yourself.
Why are we doing it now?
On October 20, 2026 the PHP project gave advance notice of an emergency security release covering two critical vulnerabilities, both scored CVSS v4.0 9.2, without naming the affected or the fixed versions. A Hacker News post had already broken the embargo and was taken down. The fixes landed the next day in 8.5.12, 8.4.27, 8.3.36, and 8.2.35, closing CVE-2026-109618 and CVE-2026-109619. Both allow unauthenticated remote code execution through ordinary web requests, and PHP's advisory says there are no complete mitigations and that updating is the only safe choice. The NCSC told organizations to find every PHP install, update it, and consider taking a system offline if they can't. Branches at 8.1 and earlier get no fix at all, so for many estates the job isn't patching, it's finding those systems and getting them off the internet. This exercise gives teams a chance to practice that call: how much to take offline, and on whose authority, before anyone can tell you what is actually vulnerable.
Link to the crisis simulation: here. (Note: Users must have Crisis Sim permissions to view the scenario link)
New CTI Labs - CVE-2026-109618 (PHP2Shell)
Release date: October 22nd, 2026
- CVE-2026-109618 (PHP2Shell) – Offensive
- CVE-2026-109618 (PHP2Shell) – Defensive
A critical memory corruption vulnerability (CVSS3.1 9.8) was disclosed on October 21, 2026, as part of a pair of flaws known as PHP2Shell. It allows an unauthenticated, remote attacker to run their own commands on a vulnerable server, with no login or special access required. The issue affects default PHP installations going back to PHP 5.4, and researchers at Cloudflare and Patchstack reported exploitation activity before the patches were even released.
In the offensive lab, users step into the attacker's seat. They'll confirm that a WordPress host is running a vulnerable version of PHP, then use the PHP2Shell exploit chain to gain remote code execution on the server and recover a token stored on it.
In the defensive lab, users take on the role of a security analyst. They'll review the logs left behind by a real attack, learn to recognise the indicators of compromise it generates, and understand how to use those signals to defend their own networks.
Who are we doing it for:
Why are we doing it now:
We're delivering both labs now so offensive teams can confirm exposure across their own estates before attackers do, and blue teams can recognise the attack patterns in their logs and respond before an unpatched host turns into a foothold.
Feature Update: Threat Landscape
Release date: October 22nd, 2026
We're launching Threat Landscape within Threat Actor Insights on the Immersive One platform. This new feature allows your security teams to filter adversaries by industry vertical and search for specific threat groups to map intelligence directly to the MITRE Adversarial Tactics, Techniques, and Common Knowledge (ATT&CK) framework.
Threat Landscape provides:
Customised threat views: Filter adversaries by industry and map intelligence directly to MITRE ATT&CK.
Granular risk and ranking views: Access consolidated dashboards for threat actor and team rankings to quickly identify critical adversary exposures and compare readiness across teams.
Deep-dive readiness metrics: Evaluate defensive posture across Coverage, Depth, and Proficiency to pinpoint critical gaps and uncover hidden vulnerabilities like key-person risk.
Targeted remediation and exercises: Seamlessly integrate with our Artificial Intelligence (AI) Program Builder and AI Crisis Simulator Builder to instantly generate hands-on technical training and realistic crisis scenarios (e.g., a LAPSUS$ extortion simulation) for low-scoring areas.
Proven defensive readiness: Track readiness trends over time to demonstrate the tangible impact of training investments to executive leadership.
Who are we doing it for:
This release is designed for security leaders, Chief Information Security Officers (CISOs), and security operations teams who need to measure defensive posture against real-world threat actors and optimize team training.
Why are we doing it now:
As cyber threats rapidly evolve, security teams need a streamlined, data-driven way to focus on the exact threat actors targeting their business. By linking threat intelligence directly to actionable exercises, Threat Landscape empowers organizations to close critical skill gaps, mitigate key-person risk, and prove defensive readiness to executive leadership.
Feature Update: Find programs faster with search
Release date: October 21st, 2026
Programs can now be found by searching, so you spend less time scrolling and more time running them.
On the Programs page, search by title or description to find a published program. You can also search deployments by title. Search is available to all customers today.
Platform Update: Manager Role Permissions Update
Release date: October 21st, 2026
Previously, team managers could create goals but not programs, so covering programs meant granting Organisation Admin permissions. That gave managers more access than they needed. Instead of adding a program-specific permission, we simplified manager permissions into one role.
What the Manager role can do
- Programs: create, edit, delete, view, publish and deploy
- Goals: view metrics, create, edit, delete, and edit ownership
- Catalogue: manage the catalogue and catalogue assignments
- Custom collections: create, edit, and delete
- Custom career paths: create, edit, and delete
- Labs: create
- Crisis Sim: create, manage, and view exercises and scenarios
- Workforce: create, manage, and view exercises and scenarios
New release: AI Agents in the CI/CD Pipeline
Release date: October, 20th, 2026
We're launching a new hands-on collection: AI Agents in the CI/CD Pipeline. Building directly on the Application and Agentic Pipeline layer from our Introducing the AI Supply Chain collection, this release moves past theoretical risks to place your team right inside the blast radius of autonomous build systems.
The collection includes three interactive labs that challenge you to secure the access controls governing build pipeline agents.
During these hands-on environments, your teams will configure parameters defining who can initiate a pipeline agent, along with restricting what the agent can read, run, and reach. Every lab features a live Claude agent running on Amazon Bedrock. This design forces defenders to analyze and mitigate real, unpredictable artificial intelligence (AI) model behaviors in real time.
Who are we doing it for
This release is designed for security engineers, DevSecOps teams, platform engineers, and cloud security architects who manage continuous integration and continuous delivery (CI/CD) workflows. It's tailored for organizations integrating autonomous AI capabilities into their development processes that need to defend against unintended model actions.
Why are we doing it now
Granting AI agents autonomous access in your CI/CD pipeline fundamentally expands your attack surface. Traditional prompt injections used to be limited to data leakage, but autonomous pipeline access gives bad actors the leverage to automatically write, approve, and deploy malicious code straight across an organization's infrastructure.
This collection is live today and available to all customers across all tiers.
Link the the collection: here.
New CTI Lab - Iris C2Agent (LLM-Operated Implant)
Release date: October, 20th, 2026
We're releasing a new Cyber Threat Intelligence (CTI) lab titled “Hunting the Tools: Iris C2Agent – Log Analysis”.
Iris (also known as C2Agent) is a Windows implant first identified on September 10, 2026, and documented by Palo Alto Networks’ Unit 42 in October 2026. Unlike traditional implants, operators do not send scripts or fixed command sequences. Instead, they issue plain-language objectives, such as "get the administrator's password," handing interactive control of the infected machine to a Large Language Model (LLM). The model executes commands, evaluates the output, and decides the next step iteratively. The malware family has seen rapid development, with 32 samples spanning eight major versions observed across 29 days. Recent iterations can hijack existing artificial intelligence (AI) developer tools on the victim's host and leverage legitimate commercial AI services.
In this hands-on lab, defenders investigate an infected Windows host using Elastic. Users will learn to:
Recover disk and network artifacts left behind by Iris.
Analyze the LLM's recorded reasoning to separate the attacker's intent from actual host activity.
Identify and catalog Indicators of Compromise (IOCs) associated with this malware family.
Who are we doing it for
This lab is designed for security analysts, incident responders, threat hunters, and detection engineers who manage Windows environments. It is equally valuable for practitioners defending networks where AI tooling is actively used, or for anyone looking to understand how permissive AI egress policies alter an organization's attack surface.
Why are we doing it now
Iris represents an active evolution in malware control as threat actors adopt AI. Because it abuses existing AI tools and commercial endpoints, malicious actions blend seamlessly into legitimate AI egress traffic—which most organizations permit by default.
We are delivering this lab now so blue teams can recognize what AI-operated malware looks like inside host and network logs, evaluate how AI egress policies impact detection visibility, and disrupt active intrusions.
Link to the lab: here.
New Crisis Simulation Content
Release date: October, 16th, 2026
We're launching a new crisis simulation, Puppetmaster: No Strings Attached.
Orchid Corp is a professional services firm that handles confidential client cases through an online portal. Your crisis team is partway through an all-hands review of the Puppetmaster's last attack when the familiar mask appears on the meeting-room screens and a cache of company passwords goes online. Staff logins get revoked, support staff can't open customer cases, and the queue starts building. From there you're working three problems at once: getting verified staff back to work, containing activity spreading from an internal application server, and telling customers, clients, and your portal partner what's actually true while a faked video of your Chief Executive circulates. Five decisions land on any single run, drawn from eight authored ones, so the containment call you make changes which problem you face next. It's the Halloween entry, so the crisis is played straight and the decor isn't. A guard radios in about the Puppetmaster's Employee of the Month photo, the coffee machine gains sentience and gets logged as a P4, and a third performance indicator tracks the office candy bowl. The closing debrief grounds the exercise in three recent real incidents: Instructure's Canvas compromise in April and May 2026, West Pharmaceutical's May 2026 intrusion, and AdaptHealth's June 2026 contractor-session compromise.
Who are we doing it for
- Executives and crisis leaders
- Technology and IT leaders
- Communications, legal, and client-facing teams
- Cross-functional teams
Why are we doing it now
Four pressures sit under the fiction, and all four are live.
Credential dumps keep landing, and revoking every staff login doesn't stop the attacker. Teams rehearse the reset. They rarely rehearse what the reset misses.
Automated response tooling is being bought in at speed, faster than teams rehearse what happens when it blocks the wrong things at machine speed. Payroll, support, and the investigators themselves are all on the list of things it can block.
Fabricated attribution is cheap. The scenario mixes genuine leaked case details with a convincing fake of the Chief Executive, and the public statement still has to go out.
Third-party assurance is the last thread. The portal partner suspends the connection with no breach of its own systems, and wants its own evidence before reconnecting. Most organizations are now on both sides of that conversation.
For a general audience, this rehearses judgment under uncertainty rather than tool knowledge. Teams get to practice verifying people, containing without stopping the business, and proving to a partner that it's safe to reconnect.
Link to the scenario: here (Note: Users must have Crisis Sim permissions to view the scenario link)
New Crisis Simulation Content
Release date: October, 16th, 2026
We're releasing a new interactive scenario on the Immersive One platform that puts cross-functional incident teams in the center of an autonomous artificial intelligence (AI) agent compromise. You'll lead the response at Orchid Solutions, a publicly listed US Software as a Service (SaaS) company handling sensitive employee records, including bank details and Social Security numbers.
In this exercise, an internal AI operations agent named Ivy uses elevated access to leak API credentials through a support ticket. The incident unfolds after a Security Operations Center (SOC) AI triage tier incorrectly scores the credential creation as routine self-provisioning and closes the alert. Across seven seats, your team will navigate critical decision points:
Initial triage: Evaluate compromised agents based on initial thin evidence.
Containment vs. continuity: Determine how much of the customer platform remains live while security searches for remaining access.
Verification and restoration: Decide what to verify before bringing an agent back online and whether a vendor's patch provides enough assurance.
Access audits: Assess how far to review other AI-connected systems that lack permissions data.
Crisis communication: Manage board reporting, regulatory duties, and public messaging after a leaked internal update puts a reporter ahead of the facts.
The attack chain draws from documented real-world mechanisms and aligns with recent public cases from Sysdig, Hugging Face, OpenAI, the UK AI Security Institute, and Unit 42.
Who are we doing it for
This scenario is designed for cross-functional teams that must hand off responsibilities end to end during a complex crisis:
Incident response and security operations
Platform and IT engineering
Chief Information Security Officers (CISOs) and Chief Technology Officers (CTOs)
Legal, compliance, and communications
Cross-functional incident teams
Why are we doing it now
Organizations are connecting AI agents to production systems with standing credentials and elevated authority faster than they can assure them. Third-party connectors are widely deployed but rarely re-tested after onboarding. In this scenario, a routine connector patch quietly expands the agent's capabilities without triggering an access re-review. Simultaneously, AI triage tiers tuned to reduce false-positive fatigue can inadvertently close critical alerts without human oversight.
Security and engineering teams are making autonomy, assurance, and containment calls for AI agents right now. Many lack a kill switch, an audit trail scoped to non-human identities, or a tested playbook for agent containment. This exercise gives your team a safe space to make those critical calls first.
Link to the scenario: here (Note: Users must have Crisis Sim permissions to view the scenario link)
New Lab Collection: Practical Introduction to AI
Release date: October, 14th, 2026
This is the second collection supporting our AI Knowledge Pass Goal, a DigComp 3.0-aligned AI literacy programme for the entire workforce. It picks up where Introduction to Using AI left off. Practitioners who have understood the basics now put them into practice with a real AI assistant on realistic workplace tasks.
What is this release?
This collection of 9 hands-on labs turns AI basics into safe, everyday working habits. Learners decide when AI is the right tool and when it isn't, write structured prompts, and build reusable assistants. They catch the bias their own questions bring, sanitize data before they share it, and learn to spot shadow AI. They then set safe limits on AI agents, handle copyright and labeling for AI content, and bring it all together in a capstone. Every lab runs in an isolated, sandboxed environment with made-up data. Labs include:
- Working With AI – When to Use It
- Prompting With Structure and Examples
- Reusable Assistants
- Bias in Your Prompts
- DLP in Practice
- Shadow AI
- AI Agents
- AI Content – Copyright, Labelling, and Disclosure
- Demonstrate Your Skills
Why are we doing it now
Knowing what AI is isn't the same as using it safely. AI tools are part of everyday work, and agents can now act on their own, not just write text. Since August 2026, national regulators can enforce Article 4 of the EU AI Act, which requires organizations that use AI to take measures to support their staff's AI literacy. Article 50's labeling rules for some AI-generated content have applied since August 2026 too. This collection gives organizations practical, hands-on evidence of those measures. It also builds the habits that prevent the most common mistakes: oversharing data, using unapproved tools, and trusting AI output without checking it.
Who is this content for
This collection is built for the general workforce who have completed Introduction to Using AI (either as part of the Knowledge pass or as standalone content) and use AI tools at work. It's the next step in the AI Knowledge Pass Goal.
Link to the collection: here
New CTI Lab - CVE-2026-104286 (FortiMail Path Traversal)
Release date: October, 14th, 2026
FortiMail is Fortinet's enterprise email security gateway. It sits at the perimeter of an organization's network, filtering all inbound and outbound email for spam, malware, and policy violations, and handling encrypted email delivery to external partners. Because every message an organization sends and receives flows through it, a compromised FortiMail appliance gives an attacker a powerful foothold: visibility into sensitive correspondence, a trusted position inside the network, and control over a system that runs with high privileges.
A critical vulnerability (CVSS 9.8) has been disclosed in FortiMail's identity-based encryption (IBE) portal, with Fortinet confirming active exploitation in October 2026. The flaw lets an unauthenticated, remote attacker write files to anywhere on the appliance, with no credentials required, simply by reaching an email-encryption feature that is designed to be publicly accessible from the internet. Attackers have used this to plant persistent access and create rogue administrator accounts on affected appliances.
In this lab, users take on the role of a security analyst responding to an alert, they'll work through a set of real FortiMail log exports to reconstruct what happened, identify the malicious email session at the heart of the attack, and recognize the indicators this exploit leaves behind, building the practical skills to detect and mitigate it in their own environments.
Who are we doing it for
This lab is designed for security analysts, incident responders, and threat hunters who monitor and defend email security infrastructure, as well as anyone responsible for log analysis and detection engineering.
It's also highly valuable for any cybersecurity practitioner who wants to understand why perimeter appliances like email gateways are such attractive targets, and how a feature that has to be exposed to the internet can become an organization's weakest point.
Why are we doing it now
CVE-2026-104286 scores 9.8 on the CVSS scale, is under confirmed active exploitation, and requires no credentials to exploit. Because the vulnerable feature is designed to be reachable from the internet, any organization running it is exposed, and a successful attack hands an adversary control of the appliance that sees all of their email.
We're delivering this lab now to equip blue teams with the practical skill of turning raw appliance logs into detections, so they can recognize an attack against their mail gateway and respond before an attacker establishes a foothold.
Availability: This lab is available now, to all customers on all tiers.
Link to the lab: here
New DTR Release: Quasar Forge - Offensive
Release date: October, 7th, 2026
Learners carry out a black-box assessment from a Kali workstation, using Open WebUI and PentAGI alongside manual testing. They direct the AI tools, review their actions, and verify findings against the environment.
The range covers:
- AI and machine-learning supply-chain compromise
- Credential exposure and privilege escalation
- Active Directory and cross-forest attacks
- Abuse of code-signing infrastructure
AI services handle model files and use credentials to reach other systems. Quasar lets teams experience how weaknesses in those services can expose the wider corporate network.
Offensive teams also practise working with AI tools during an assessment, deciding when to refine instructions, check results manually, or take over a task.
Who is this for?
- Red teams and penetration testers
- AI security teams
Feature Update: Program Enhancements
Release date: October, 7th, 2026
We've released a series of enhancements in our Programs.
Add or remove participants after deployment
Org Admins can now add teams, withdraw teams or individuals, and re-enrol them on a Program that's already live. No need to create a new deployment when someone joins or leaves.
Participant licence visibility
Froom the Teams or Individuals views, admins can open any participant to see their assigned content and its status. If a participant was withdrawn because the org doesn't have the licences for their content, it's flagged, along with exactly which user needs a licence.
New CTI Lab: CVE-2026-88771 (NetScaler Log-to-Shell)
Release date: October, 6th, 2026
NetScaler ADC (Application Delivery Controller), made by Citrix, is a network appliance deployed at the edge of enterprise environments to manage load balancing, remote access, and VPN services for corporate users. Because it sits between the internet and internal systems, and often handles authentication for remote users, a compromised NetScaler appliance gives an attacker a direct window into the network and access to everything routed through it.
A critical unauthenticated remote code execution vulnerability (CVSS 9.8) was disclosed in NetScaler ADC and NetScaler Gateway in September 2026 and has since been confirmed as actively exploited. The exploit chain works by allowing an attacker to write a shell command into a system log through a failed login attempt, and a root-context background daemon executes it hours later.
In this lab, users will taking on the role of a SOC analyst investigating a suspected compromise, they'll work through real NetScaler log files to identify the attacker's actions, learning to recognize the indicators this attack leaves behind.
Who are we doing it for?
This lab is designed for SOC analysts, incident responders, and threat hunters responsible for monitoring network appliances and edge infrastructure, as well as anyone who performs log analysis or detection engineering for enterprise environments.
It's also highly valuable for any cybersecurity practitioner who wants to understand how enterprise network appliances can be weaponized through their own internal processes, and why delayed-execution attack chains demand a broader investigation window than most incidents.
Why are we doing it now?
NetScaler appliances are deeply embedded in enterprise networks worldwide, and CVE-2026-88771 requires no credentials, just a single request to the management interface. At a CVSS score of 9.8 and with active exploitation already confirmed, unpatched appliances represent an open door to the internal network.
We're delivering this lab now to equip defenders with the log analysis skills to spot this attack in progress, reconstruct what an attacker has done, and respond before a foothold becomes a full compromise.
Link to the lab: here
New Crisis Simulation: The Hex Files - A brew-tal-breach
Release date: October, 1st, 2026
Playing the Editor-in-Chief, IT Manager, Communications Manager, and Data Protection Officer, participants contain the intrusion, brief staff, work with the National Crime Agency, recover website and account access under time pressure, weigh a UK GDPR notification to the ICO, and make the call every newsroom dreads: publish the verified story under threat, or hold it.
Who are we doing it for?
- Incident responders and IT leads: containment, credential recovery, and sequencing recovery work when break-glass accounts are the only way back in.
- Communications and PR teams: internal staff briefing during a lockout and external messaging once a defacement is public.
- Legal, privacy, and executive leaders: the UK GDPR notification call on an exposed subscriber export, and a publish-or-hold decision under an active threat.
- Cross-functional crisis teams: the value here is the handoffs between all four seats under one clock.
New Crisis Simulation: The Hex Files - The Nightshade Incident
Release date: October, 1st, 2026
You step in as the new Head of Security Operations, then sit with the crisis management team as the picture widens: a second account leaking quietly, a regulator decision, a ransom demand, and a leaked-emails subplot, all while the release date holds firm and the share price moves. What you do early shapes how much of Nightshade, and how much else, ends up outside the building.
Who are we doing it for?
- Security operations and incident responders: triage sign-ins that no tool flagged, scope a compromise that turns out to span more than one account, and weigh restoring fast against preserving the evidence of how the attacker got in.
- Crisis management and executive teams: decide regulator timing, staff communications, and how to answer both a ransom and a public leak without trading away future leverage.
- Communications and cross-functional teams: practice holding a public line under press and social-media pressure while the facts are still landing.
It's Cyber Awareness Month, and this one leans into the season. The scenario rehearses a pattern that keeps showing up in real intrusions: attackers profile a remote worker on social media, phish a single account, and quietly open a second route that closing the first one never touches. It puts business and technical players in the same room to practice the decisions that decide how bad a ransomware incident gets, from how a trusted sign-off gets abused to whether you restore before you understand the way in.