Artificial Intelligence: Navigating the Evolving Landscape
The changing world To understand where we're going, you first need to grasp the sheer scale of what's happening now. The May 2025 report on Artificial Intelligence Trends by Mary Meeker and Bond Capital paints a vivid picture of a sector in overdrive: Unprecedented user adoption: Generative AI tools have achieved mass adoption faster than any previous technology, including the internet and smartphones. Soaring infrastructure investment: Top tech giants (Apple, NVIDIA, Microsoft, Alphabet, Amazon, Meta) spent a combined $212 billion on capital expenditures in 2024, a huge portion of which was dedicated to AI infrastructure like data centres and custom silicon. Shifting cost dynamics: The cost to train a state-of-the-art foundation model remains astronomically high, somewhere in the hundreds of millions of dollars. However, the cost to use these models (the inference cost) is plummeting, making AI more accessible than ever before. Intense competition and rapid imitation: AI is boosting productivity and driving competition between products. Global AI "space race": Nations are treating AI supremacy as a strategic imperative, leading to significant government investment and policy-making, particularly in areas like the semiconductor supply chain, with the US, Europe, and China all building new fabrication plants. With this level of investment and adoption, can you confidently say this is a bubble about to burst? Sir Demis Hassabis, CEO of Google DeepMind, puts this huge change on the same magnitude as the industrial revolution and the launch of the internet. Data from Gartner supports this, suggesting that by the end of 2025, 39% of organizations worldwide will have moved into the experimentation phase of AI adoption. The shift is well and truly on. What does AI look like in 2025? AI is underpinned by machine learning models, which are trained, not programmed. Engineers feed them vast amounts of data, and they learn patterns, concepts, and relationships. Different types of models are used for different purposes, such as those specialising in human language interactions (large language models, LLMs) and artwork generation (diffusion models). When using AI systems, such as chatbots, you’re not interacting with the model directly but rather with additional software that uses the model as its “brain”. This allows you to implement guardrails to check user inputs and model outputs, helping to filter out harmful or inappropriate content. Modern AI systems are rarely just a wrapper around a model. They integrate with other tools and services to enhance their capabilities, such as searching the web for real-time information or accessing private company documents to provide context-specific answers. The year of agentic AI An AI agent is a system that can autonomously pursue a goal. Instead of responding to a single prompt, it can reason, plan, and execute a series of steps to accomplish a complex task. It can also decide which tools to use and in what order. An AI agent may still be a chatbot or run constantly in the background. Big tech companies are adamant that agentic AI is the next evolution, with Google, Amazon, and Microsoft all predicting the next wave of innovation over the next two years. A key catalyst for this explosion was the release of the open-source Model Context Protocol (MCP) by Anthropic in late 2024. MCP provides a standardized way for AI models to discover and use tools. As the official documentation puts it: "Think of MCP like a USB-C port for AI applications. Just as USB-C provides a standardized way to connect your devices to various peripherals... MCP provides a standardized way to connect AI models to different data sources and tools." Source: Model Context Protocol - Getting Started MCP has been a game-changer, dramatically simplifying the process of giving AI systems new capabilities and accelerating the move from AI systems that know things to AI systems that do things. It’s no coincidence that technology companies then started to release their guides for building AI agents following MCP’s release – with Microsoft, Google, Cloudflare, OpenAI, and Amazon following close behind. Technology to watch Finally, a few key technologies that will define the next phase of AI include: Model Context Protocol (MCP) Continue to watch this standard. As more tools and platforms adopt MCP, the ecosystem of "plug-and-play" capabilities for agents will explode, as will the security risks. Simon Willison puts it perfectly by describing a “lethal trifecta”. AI systems with access to private data, the ability to communicate externally, and exposure to untrusted content could easily lead to serious consequences. Source: Simon Willison Authorisation for AI systems As agents move from knowing things to doing things (e.g., booking travel, purchasing supplies, modifying code), security becomes paramount. We need robust authorisation. This will involve human-in-the-loop (HITL) approvals, likely powered by modern authentication standards like Client-Initiated Backchannel Authentication (CIBA), which can send a push notification to a manager to approve an agent's action. Thought leaders from Microsoft suggest an overhaul to OAuth, with agentic systems having their own distinct identities and security considerations. One thing’s for sure: proper authorization is complex – difficult to get right and catastrophic to get wrong. Agent-to-agent communication Current AI agents are specialized for a specific purpose, but next-generation AI functionality comes through the use of multi-agent systems, which can be deployed in a variety of architectures, such as hierarchical or swarms. How agents communicate with each other, share memory, and share capabilities is still in its relative infancy, especially when AI agents may be hosted independently and written with different frameworks. Two competing protocols are emerging: Google's Agent2Agent protocol and IBM’s Agent Communication Protocol (ACP). It's too early to call a winner, but the development of a standard here will be a major milestone. We are at the beginning of the agentic era. 2025 is the year for experimentation. It's time to move from simply using AI to actively building with it, automating the tedious, and unlocking new forms of creativity and productivity. Getting the most out of AI If one thing’s for sure, it’s that the AI landscape is moving fast. So it’s crucial that you and your organisation are at the forefront of AI developments and making the most out of the latest technologies. Keep your eyes peeled for brand new labs in this space coming very soon! Our brand new collection will demystify terminology, explore the core concepts, and let you build and secure modern AI systems in a safe, sandbox environment. Sign up for email notifications from the Immersive Community so you don’t miss out on this brand new collection.New CTI Labs: CVE-2025-53770 (ToolShell SharePoint RCE): Offensive and Defensive
Recently, a critical zero-day vulnerability affecting on-premise SharePoint servers, identified as CVE-2025-53770, was uncovered. This vulnerability allows for authentication bypass, leading to remote code execution, and has been actively exploited in the wild. Eye Security researchers detected an in-the-wild exploit chain on July 18, 2025, during an incident response engagement. This discovery led to Microsoft assigning two CVEs: CVE-2025-53770 and CVE-2025-53771. The attack notably leveraged a combination of vulnerabilities to achieve its objectives, impacting numerous SharePoint servers globally. There is now a public exploit available for anyone wanting to achieve remote code execution. Why should our customers care? This critical vulnerability has been added to the CISA Kev Catalog. and with no authentication or user interaction, a vulnerable SharePoint server can be fully taken over remotely, letting attackers run arbitrary code as if they were privileged admins. SharePoint is a complex and large system that often holds a lot of sensitive data for organizations and is often a targeted system for attackers. Who is the defensive lab for? System Administrators SOC Analysts Incident Responders Threat Hunters Who is the offensive lab for? Red teamers Penetration Testers Threat Hunters Here are the links to the labs: Offensive: https://immersivelabs.online/v2/labs/cve-2025-53770-toolshell-sharepoint-rce-offensive Defensive: https://immersivelabs.online/v2/labs/cve-2025-53770-toolshell-sharepoint-rce-defensiveNew CTI Lab: CVE-2025-32463 (Sudo Chroot Elevation of Privilege): Offensive
On June 30, 2025, the Stratascale Cyber Research Unit (CRU) team identified a critical local privilege escalation vulnerability in sudo, tracked as CVE-2025-32463. This vulnerability, related to sudo's chroot option, can allow an attacker to escalate privileges to root on an affected system. Why should our customers care? This critical vulnerability is reasonably trivial to exploit, and should an attacker gain user-level access to a vulnerable machine, they'll be able to elevate their privileges and have full control over the machine. It has come to our attention that not many people are aware that sudo has versioning. It is a binary that is constantly iterated upon, which naturally may introduce new vulnerabilities. If administrators and security analysts are not aware of how these vulnerabilities work, this can lead to significant risks and impacts. Who is it for? Red Teamers Penetration Testers System Administrators Here is a link to the lab: https://iml.immersivelabs.online/labs/cve-2025-32463-sudo-chroot-elevation-of-privilege-offensiveNew CTI/OT Lab: Norwegian Dam Compromise: Campaign Analysis
We have received reports of a cyber incident that occurred at the Lake Risevatnet Dam, near Svelgen, Norway, in April 2025. A threat actor gained unauthorized access to a web-accessible Human-Machine Interface (HMI) and fully opened a water valve at the facility. This resulted in an excess discharge of 497 liters per second above the mandated minimum water flow. Which persisted for four hours before detection. This attack highlights a dangerous reality: critical OT systems are increasingly exposed to the internet, making them accessible to threat actors. In this case, control over a dam’s valve system was obtained via an insecure web interface, a scenario that could have had even more severe consequences. A recent report by Censys identified over 400 exposed web-based interfaces across U.S. water utilities alone. This dam incident in Norway exemplifies the tangible risks posed by such exposures. In this lab, you will be taken through the attack from an offensive viewpoint, including cracking an HMI and fully opening two valves. Why should our customers care? OT environments, including dams, energy grids, and oil pipelines, are foundational to national security and daily life. These systems cannot be secured using traditional IT playbooks. As OT becomes more connected, tailored security strategies are critical to prevent unauthorized access and catastrophic failures. Who is it for? Incident responders SOC analyst Threat Hunters Red Teamer Penetration Testers OT Engineers Here is the link to the lab: https://immersivelabs.online/v2/labs/norwegian-dam-compromise-campaign-analysisNew CTI Lab: CVE-2025-33073 (SMB Elevation of Privilege): Defensive
Another vulnerability patched was released during Microsoft's June 2025 patch Tuesday review! An important elevation of privilege vulnerability was listed, and if exploited successfully, attackers can achieve elevation of privilege on the compromised machine. Even though it's not recorded to have been exploited in the wild as yet, the fact that research exists with details on how the vulnerability was found improves the chances an attacker will attempt to exploit this flaw against a victim.In these labs, you will be taken through the vulnerability from both an offensive and defensive perspective. Why should our customers care? This is a new vulnerability that has just been patched, and is has in depth research released about it. Successful exploitation of this vulnerability allows attackers to elevate their privileges and achieve command execution on a victim machine. Learn what sort of indicators this exploit leaves, but also learn how to execute and take advantage of this vulnerability! Who is it for? Incident responders SOC analyst Threat Hunters Red Teamer Penetration Testers Here is the link to the labs: Defensive: https://immersivelabs.com/v2/labs/cve-2025-33073-smb-elevation-of-privilege-defensive Offensive: https://immersivelabs.com/v2/labs/cve-2025-33073-smb-elevation-of-privilege-offensive Container 7 Release We have released a threat detection for this particular vulnerability, helping the community to protect against any potential use of this vulnerability. https://github.com/Immersive-Labs-Sec/SigmaRules/blob/main/cve-2025-33073-smb-exploit.ymlNew CTI Lab: Stealth Falcon (CVE-2025-33053) – WebDAV Server Remote Code Execution
Yesterday, in Microsoft's Patch Tuesday, there was a zero-day vulnerability that was patched and has been exploited in the wild! This zero-day was used by the cyber-espionage group Stealth Falcon and was reported on by Checkpoint. After successful phishing attempts, the user will execute a .url file that exploits a vulnerability to communicate with a WebDAV server owned by attackers, which holds a particular binary. The vulnerability is present because Windows will look for binaries through the WebDAV link before searching for the legitimate one on its own PC. Therefore, the attackers can achieve remote code execution. We are releasing a lab on hunting for the execution of this vulnerability to help teams create effective threat detections. Why should our customers care? This is a new vulnerability that has just been patched and has already been successfully used as part of threat groups' attack chains. Therefore, it is recommended to see what sort of indicators of compromise this type of vulnerability leaves once exploited. Who is it for? Incident responders SOC analyst CTI Analysts Threat Hunters Here is the link to the lab: https://immersivelabs.online/labs/stealth-falcon-cve-2025-33053-webdav-server-exploitation As part of the Container 7 team, we have also released threat detections that cover both binaries that were used in the campaigns that exploited CVE-2025-33053. You can find these here: https://github.com/Immersive-Labs-Sec/SigmaRules/blob/main/cve-2025-33053-iediagcmd-exploit.yml https://github.com/Immersive-Labs-Sec/SigmaRules/blob/main/cve-2025-33053-CustomShellHost-exploit.ymlNew CTI Labs: BadSuccessor: Offensive and Defensive
Two days ago, Akamai released a technical research blog post detailing a privilege escalation vulnerability in Windows Server 2025. This vulnerability abuses delegated Managed Service Accounts (dMSAs), and with the right base permissions, it could allow a user to gain domain admin permissions or even dump the NTLM hashes for all users in the domain. There is no patch available, and this would be considered a public zero-day. Why are these labs important? Many organisations use a Windows Domain to manage their users and accounts. This newly announced zero-day has no patch and no known detections in SIEMs. A combination of these labs will allow organisations to identify any potentially weak configurations vulnerable to exploitation and how to threat hunt in a SIEM to identify signs of exploitation. Who is it for? Incident responders SOC analyst CTI Analysts Threat Hunters Pentesters / Red Teams Here is the link to the analysis lab: BadSuccessor – Offensive BadSuccessor – DefensiveNew CTI Lab: Sandworm Campaign: ZEROLOT Wiper
ESET released a new APT threat report today, and amongst the information was a new malware wiper used to attack critical national infrastructure. However, this malware has not been reported on at all. It has been successfully deployed amongst many organizations, but no analysis has been released. Therefore, we are releasing a SIEM analysis to help our customers create threat detections for this destructive malware. The threat actor in question is Sandworm Team, a state-sponsored APT group that has been active since at least 2009. Known for highly destructive cyber campaigns, the group has targeted critical infrastructure. In this lab, you'll be exposed to one of Sandworm's latest campaigns, where they use remote management tools to facilitate the deployment of a new wiper, Zerolot. Why is this lab important? Many of our customers have asked for an analysis of wiper malware, and the destructive nature of this malware worries organizations around the world. This new strain, which has been deployed numerous times successfully since December 2024, needs effective threat detection to ensure security teams are prepared for this threat. Who is it for? Incident responders SOC analyst CTI Analysts Threat Hunters Here is the link to the analysis lab: Sandworm Campaign: ZEROLOT WiperTransforming Bug Triage into Training: Inside the Making of Immersive AppSec Range Exercises
“We all know the pain of bug reports clogging up a sprint—we thought, what if we could transform that drain on time and morale into a challenge developers are excited to tackle?” Rebecca: Oh, I love that—turning bug backlog dread into bite-sized victories is brilliant. I’m excited to hear more, but first, congratulations on launching Immersive AppSec Range Exercises! This is a BIG deal! No one else does anything like this for developers. Naomi: Thanks! What can I say? My love for cybersecurity goes back to university capture-the-flag events. Pushing yourself outside your comfort zone with hands-on challenges is by far the fastest way to learn. My main goal was to bring that same energy to application security—there are loads of CTFs for pentesters, but not really for developers who need to sharpen their defensive and remediation skills. I also wanted this to be inherently team-friendly. Our individual AppSec labs are built for individual learning, but group dynamics demand different pacing and collaboration tools. Rebecca: Makes total sense. Offensive skills get the headlines, but developers need a solid, team-centric defensive playground too. So how did you translate that vision into the actual structure of our AppSec Range Exercises? Naomi: I anchored everything in the maintenance phase of the software lifecycle: Receive bug → Triage → Fix → Test → Merge. That mirrors real dev workflows, so participants don’t just patch vulnerabilities—they live the ticket management, version control, and testing cadence they’ll face on the job. [Inside scoop: When we build any security exercise, our team maps it to a real-world experience. In Immersive AppSec Range Exercises, a common SDLC workflow—teams learn best when they see exactly how it will play out in their daily sprints. ] Rebecca: I love that you’re training both mindset and muscle memory—jumping through the same process you’d use in production. Once you had that flow, what were the first steps to bring the framework to life? Naomi: Well, I knew that this project was going to need quite a few applications to house the functionality for the exercises, so I audited what we’d need from scratch versus what open source could handle. For ticketing, most OSS Kanban tools were overkill, so I built a lightweight app called Sprinter. Then for version control, we leaned on GitLab—it was quick to stand up and gave a familiar UI for branching and merges. Once those pieces clicked—vulnerabilities surfacing in Sprinter, code pushes in GitLab, and test runs in the Verification view—we had a minimally viable range exercise in action. Rebecca: A smart “build-what-you-must, borrow-where-you-can” approach. Seeing that prototype come together must’ve been so cool. Naomi: Absolutely. It was one thing to design on paper, but watching the pipeline live—tickets flow in Sprinter, GitLab merge requests, automatic test feedback—was a genuine “wow” moment. Rebecca: Speaking of “wow,” let’s talk scenarios. How did you land on “Blossom,” your vulnerable HR app in the Orchid Corp universe? Naomi: Well, we needed something with enough complexity to showcase the framework. HR apps hit three sweet spots: business logic richness, varied user roles, and sensitive data. Tying it into Orchid Corp—our fictional corporation for Immersive Cyber Drills—gave it narrative depth, especially for returning users of our Immersive One platform. Rebecca: And when you designed the actual vulnerabilities inside Blossom, what guided your choices? Naomi: I started with the OWASP API Top 10—that’s our gold standard for spotting the biggest threats. Then I looked at what slips through most scanners and frameworks—nuanced business-logic flaws and edge-case logic bugs—and made those the core of the challenge. To keep things well-rounded, I also added a few classics—things like IDOR, SSRF, and command injection—so every player gets a taste of both modern pitfalls and time-tested exploits. [Inside scoop: Mixing modern, real-world API flaws with a few known “gotchas” keeps Immersive AppSec learners guessing and builds confidence when they spot the unexpected.] Rebecca: I know you’re busy working on the next exercises we’ll release, but before we wrap, how did you test Blossom among developers and engineers? No doubt you wanted to make sure it delivered the right experience! Naomi: Yes, absolutely! We ran a pilot with our own Immersive engineers and a third party, creating a realistic dev team. Watching them collaborate—triaging, patching, merging—validated every piece of the design. Their feedback on pacing and hint levels let us polish the final release. It was one of my favourite days—seeing months of work click into place. After that, we shipped it to customers knowing it was battle-tested. Rebecca: This has been fantastic—thank you for sharing your full planning and development journey, Naomi! From initial vision to a live, collaborative exercise … I’m awed. You certainly put incredible thought and care into developing this revolutionary approach to AppSec training. Final Thought Security is a team sport, and training like Immersive AppSec Range Exercises is the fast track to confident, resilient DevSecOps teams. If you’re a developer or engineer looking to level up your remediation skills, have your team lead reach out to your Account Manager for a demo. In the meantime, watch a sneak peek of what your experience would be like in this demo below:New CTI Labs: Threat Actors Akira and DragonForce
These labs will highlight the background and TTPs of Akira, a highly prolific threat actor with indiscriminate targeting, and DragonForce, a ransomware actor recently in the news, connected to the attacks on M&S, Co-op, and Harrods. Why are these labs important? Akira is one of the most prolific threat actors that does not discriminate in its targeting. It often targets medium to large enterprises worldwide, with a strong focus on North America and Europe, including the UK. This means that no one is exempt from Akira's targeting in the future, so knowing your TTPs and how to prepare for attacks from threat actors is paramount to keeping your organization safe. Throughout late April 2025, DragonForce has been all over the news since they claimed responsibility for being involved in the attacks against Marks and Spencer, Co-Op, and Harrods in the UK. The information in the labs reflects DragonForce's latest known TTPs and helps customers to stay one step ahead of actors like DragonForce. Who are these labs for? These labs deliver the latest information relating to threat actors and their TTPs. The personas who would benefit the most from these labs are: Cyber Threat Intelligence Analysts SOC Analysts Incident Responders Threat Hunters Here are the links to the labs: Threat Actors: Akira Ransomware Groups: DragonForce
