Deep Security / Server & Workload Protection

Trend Deep Security (the vendor's enterprise security business unit has been known as TrendAI™ since March 23, 2026; the Deep Security product name has not changed) is a hybrid cloud workload security platform that protects physical servers, virtual machines on VMware and Hyper-V, and cloud workloads on AWS, Azure and Google Cloud with a single agent and a single console. It brings together server-specific controls such as host-based intrusion prevention (IPS), virtual patching, file integrity monitoring (FIM) and log inspection; the aim is to prevent a server from being exploited during the window that stays open until a patch is applied.

That window is a measurable risk. According to the Verizon 2025 Data Breach Investigations Report, vulnerability exploitation as an initial access vector grew 34% in a single year and now accounts for 20% of breaches; the same report shows that only about 54% of edge device vulnerabilities were fully remediated, and that the work took a median of 32 days (Verizon DBIR, 2025). Although that measurement concerns edge devices, the mechanism is the same on the server side: the delay between a patch being released and actually being deployable is the interval in which the attacker operates. Deep Security does not close the vulnerability in that interval — it closes the exploitation path.

Which protection modules does Trend Deep Security include?

The official Deep Security 20 documentation defines eight protection modules: Intrusion Prevention, Anti-Malware, Firewall, Web Reputation, Integrity Monitoring, Log Inspection, Application Control and Device Control (Deep Security 20 Help Center, 2024). All modules run on the same agent; which module is enabled on which server is decided at the policy level. A web server and a database server can therefore be managed from the same console with independent rule sets.

• Intrusion Prevention (host IPS): Shields known vulnerabilities with rules until the vendor patch is applied. The official documentation calls this function virtual patching outright.
• Anti-Malware: Detects malicious software on the server using signatures, behavior and machine learning.
• Firewall: Provides stateful packet filtering per server, isolating the workload independently of the network segment it sits in.
• Web Reputation: Performs reputation checks on outbound connections from the server, making it harder for command-and-control traffic to get out.
• Integrity Monitoring (FIM): Monitors all authorized and unauthorized changes to files, directories, registry entries and services, and raises alerts.
• Log Inspection: Collects and analyzes system logs and produces audit evidence for PCI DSS or internal requirements.
• Application Control: Inventories the executables on the server and blocks unapproved software from running.
• Device Control: Restricts removable media and peripherals attached to servers.

This distinction matters in practice: integrity monitoring and log inspection produce audit evidence that a classic server antivirus cannot generate. According to the official documentation, Integrity Monitoring records not only attacker-driven changes but authorized ones as well. That allows the change management process and security telemetry to be compared against the same record set; an unplanned intervention is noticed the moment it fails to match the planned maintenance records.

What is virtual patching, and does it replace a real patch?

No — virtual patching does not replace the patch; it makes the patch schedule achievable. The intrusion prevention module blocks, at rule level, the network pattern or call path through which a vulnerability is exploited; the vulnerability remains on the server but becomes non-exploitable. For systems with no available downtime window, servers that cannot be updated because of application compatibility, and out-of-support operating systems for which the vendor no longer issues patches, this is often the only technical control that can be applied.

The business rationale comes directly from compliance regulation. PCI DSS v4.0 Requirement 6.3.3 mandates that critical or high-severity security patches be installed within one month of release (PCI Security Standards Council, 2022). For organizations that cannot meet the one-month window through change management alone, virtual patching is the interim control that brings risk down to a bearable level during that window.

The intelligence source behind the rule set is Trend's Zero Day Initiative (ZDI) program. ZDI states that it has coordinated the disclosure of more than 15,000 vulnerabilities since 2007; the program is fed by research teams in 14 global threat centers as well as more than 19,000 independent researchers (Zero Day Initiative, 2026). According to Omdia's "Quantifying the Public Vulnerability Market: 2025 Edition" research, Trend single-handedly reported 73% of the vulnerabilities disclosed globally in calendar year 2024 (Zero Day Initiative / Omdia Research, 2024 data).

The critical detail is in ZDI's disclosure policy: the policy grants the vendor a standard 120-day period for a patch, but explicitly states that protection filters may be distributed to Trend customers at the same time as the vendor is notified (ZDI Disclosure Policy, 2026). That is the official basis for virtual patching taking effect before the vendor patch. Thanks to the same intelligence chain, TippingPoint network IPS customers were protected an average of 96 days ahead of the vendor patch across more than 1,211 vulnerabilities disclosed in 2023, according to the official datasheet (TippingPoint TXE Series datasheet, 2025).

The ransomware counterpart of that head start has been measured as well. According to Trend's 2025 Cyber Risk Report, 59 zero-day vulnerabilities have been used in ransomware attacks since 2020; before 2020 it was extremely rare for ransomware groups to use a zero-day (Trend 2025 Cyber Risk Report). Zero-day exploitation entering the ransomware economy turns protection arriving ahead of the vendor patch from an optional improvement into a baseline requirement.

Which virtualization and cloud environments does Deep Security run in?

Deep Security runs both with an agent and — under certain conditions — agentless. When the Deep Security Virtual Appliance is used in a VMware NSX environment, the anti-malware, integrity monitoring, firewall, intrusion prevention and web reputation modules can run agentless; in the older vShield environment, agentless coverage is limited to anti-malware and integrity monitoring only (Deep Security 20 Help Center, 2024). In data centers hosting large numbers of virtual machines, this is how protection is delivered without installing an agent on every VM; it also removes the burden of managing agent lifecycles on short-lived machines built from templates.

On the cloud side, the Deep Security Agent can be installed with full support on supported platforms running on Amazon AWS, Google Cloud Platform and Microsoft Azure (Deep Security Agent platforms, 2024). Inventory management is automated through connectors: when an AWS account is added to Deep Security Manager, all EC2 and WorkSpaces instances are imported into the console; when a GCP account is added, all GCP VM instances are imported; the Azure connector synchronizes the VM inventory every 10 minutes (Deep Security 20 Help Center, 2024). In auto-scaling cloud environments, keeping a newly launched server from running without a policy depends on this mechanism.

On containers we do not present the scope as broader than it is. The official agent compatibility documentation defines Docker host and container support on Linux distributions. Cluster-level Kubernetes protection — registry image scanning, admission control, pod-level policy — is not declared in the current Deep Security 20 LTS documentation, so we position that scope as a separate component, with the container security module on the Trend Vision One platform side. If you have an architecture running on EKS, AKS or GKE, we clarify the scope in writing at the start of the project.

What does Deep Security deliver in PCI DSS and KVKK audits?

Deep Security's compliance value is not in closing a single requirement but in producing both a technical control and auditable evidence for several requirements at once. PCI DSS v4.0 Requirement 11.5.1 mandates that intrusion detection and/or prevention techniques monitor all traffic at the perimeter of the cardholder data environment (CDE) and at critical points within it, and that signatures and baselines be kept current; Requirement 11.5.1.1 has made the detection of covert malware communication channels mandatory for service providers as of March 31, 2025. Requirement 11.5.2 mandates deploying a change-detection mechanism such as file integrity monitoring and performing critical file comparisons at least once per week (PCI Security Standards Council, 2022).

A misunderstanding we encounter frequently should also be corrected: virtual patching is not an "approved compensating control" in PCI DSS. The PCI SSC does not approve any specific technology as a compensating control in any of its documents; Appendix B of the standard only states that organizations with a legitimate and documented technical or business constraint may take that route by completing a Compensating Controls Worksheet (CCW), and the acceptance decision rests entirely with the QSA's assessment. The correct framing is this: intrusion prevention is already a direct requirement of 11.5.1, and file integrity monitoring of 11.5.2. The compensating control discussion only arises when the one-month patch window in 6.3.3 cannot be met — and even then, the organization itself must document the justification.

On the KVKK (Turkey's data protection law) side, the basis is Article 12 of Law No. 6698: the data controller is obliged to take all technical and administrative measures needed to ensure an appropriate level of security, in order to prevent unlawful processing of personal data, prevent unlawful access to it and ensure its preservation (KVKK, 2016). Table 4.1, the Summary Table of Technical Measures in the Authority's Personal Data Security Guide, explicitly lists intrusion detection and prevention systems, log records, firewalls and up-to-date anti-virus systems; the same guide identifies patch management and software updates as one of the elements needed to close potential vulnerabilities (KVKK Personal Data Security Guide). These are the items Deep Security covers at the server layer.

The guide also recommends layered and regularly reviewed complementary measures, stating that "the view that full security can be achieved through the use of a single cyber security product is not always correct". For that reason we position Deep Security not on its own but together with the endpoint, network and email layers, within the whole of the Trend Micro solution family. There is similar regulatory pressure in Europe: according to ENISA Threat Landscape 2025, 53.7% of the incidents analyzed in the EU concern essential entities within the scope of the NIS2 Directive, and the most targeted sector is public administration at 38.2% (ENISA, 2025).

What is the difference between Deep Security, Apex One and TippingPoint?

The three protect different layers and do not replace one another. Deep Security operates at the server and workload layer, Apex One endpoint protection at the user device layer, and TippingPoint inline in the network segment. The typical problem with using an endpoint agent on servers is that server-specific controls — integrity monitoring, log inspection, server-based firewall and virtual patching — are missing; Deep Security fills exactly that gap.

If deep visibility is needed at the network layer, the architecture is completed with Deep Discovery network detection and sandbox analysis; in organizations running the perimeter firewall layer with a different vendor, a design that works alongside the existing firewall can be built. Deep Security's position on the host also lets it record in-server events that a network device cannot see in east-west traffic — with encrypted internal traffic, that difference is decisive.

How does Deep Security combine with Trend Vision One XDR?

Deep Security can run independently with its own console, but its value increases when its telemetry is fed into XDR correlation. Trend Vision One has six native security sensors: endpoint, cloud, email, network, server and identity (Trend Micro Newsroom, 2025). Thanks to the server sensor, a chain that starts with a phishing email and jumps from the endpoint to the server can be seen in a single incident graph. Otherwise the IPS block on the server and the suspicious process on the endpoint remain two separate alerts, examined individually and often never correlated.

On the log side, the Agentic SIEM component of Vision One became available on August 1, 2025; it supports more than 900 data sources and offers up to 2 years of analytical retention and up to 7 years of archive retention (Trend Micro Newsroom, 2025). In organizations that must keep long-term audit evidence, where the Deep Security log inspection output will be written — the organization's own SIEM or the Vision One side — should be decided as an architectural matter at the start of the project. Another reason to keep an independent audit trail at the server layer is supply chain risk: according to the Verizon 2025 DBIR, the share of breaches involving a third party doubled in a single year, rising from 15% to 30% (Verizon DBIR, 2025). Having the changes made by an external supplier accessing the server for maintenance or support show up in integrity monitoring and log inspection records is the basis of any post-incident review.

How does Trend Deep Security licensing work and where do you buy it?

Licensing is structured around the number of servers and workloads protected, the module set enabled and the contract term. An on-premises Deep Security deployment and a cloud-side Workload Security subscription are offered under different commercial models. We do not publish prices on this page; an accurate figure only emerges once the inventory (physical servers, virtual machines, cloud instances), module scope and support level are clear. Let us take stock of your environment together and request a proposal — a figure given without measurement would be misleading.

As an authorized Trend Micro channel partner, Sora Yazılım provides licensing, deployment, KVKK-compliant management and incident response together. On the vendor side, global support capacity rests on 6,000 TrendAI experts across 75 countries (Trend Micro Newsroom, 2026). Because product naming was updated with the March 2026 brand transition, we confirm in writing which SKU covers which module in contracts; module scope should never come as a surprise later.

How does the rollout process work?

We run the process in four steps. In the first step, inventory and scope are established: which servers are in the cardholder data environment, which run out-of-support operating systems, and which cannot be patched because of application certification. In the second step, agents are installed on a limited pilot group and IPS rules are run in detect mode only at first; no rule is switched to blocking mode until it has been verified not to generate false positives in production traffic. In the third step, policy sets are separated by server role. In the fourth step, log inspection output is forwarded to the SIEM and audit reporting is automated.

In cloud and hybrid architectures we carry out the deployment with automation as part of our DevOps and infrastructure services: agent installation is embedded into the machine image or into configuration management tools, so a newly launched server in an auto-scaling group never runs without a policy. In projects that require hardware refresh, we proceed together with capacity planning on the HP server solutions side. To strengthen the recovery side in a ransomware scenario, we recommend an integrated plan with Acronis Cyber Protect backup; the protection layer and the recovery layer must be tested separately.

In summary: Trend Deep Security makes the patch window manageable on server workloads with host IPS and virtual patching, produces evidence for PCI DSS and KVKK audits with file integrity monitoring and log inspection, and applies the same policy logic agentless on VMware NSX and agent-based on AWS/Azure/GCP. Let us assess your server inventory, compliance scope and unpatchable systems together and draw up a module-level scope and license plan: request a proposal through the contact form and let our technical team start with the inventory exercise.