Blog
/
Network
/
May 19, 2026

When Open Source Is Weaponized: Analysis of a Trojanized 7 Zip Installer

Attackers abused a trojanized 7‑Zip installer distributed via a typo‑squatted domain to establish proxyware infections worldwide. This blog analyzes how social engineering, trusted open‑source software, and stealthy command‑and‑control activity enabled widespread compromise, highlighting the importance of software validation, user awareness, and proactive network‑level detection.
Inside the SOC
Darktrace cyber analysts are world-class experts in threat intelligence, threat hunting and incident response, and provide 24/7 SOC support to thousands of Darktrace customers around the globe. Inside the SOC is exclusively authored by these experts, providing analysis of cyber incidents and threat trends, based on real-world experience in the field.
Written by
Justin Torres
Cyber Analyst
7 zip installerDefault blog imageDefault blog imageDefault blog imageDefault blog imageDefault blog imageDefault blog image
19
May 2026

Background of the malicious 7-Zip installer, and assessing its Impact

Early in 2026, external researchers disclosed a malicious distribution campaign leveraging a trojanized installer masquerading itself as a legitimate 7‑Zip utility. Evidence suggests the campaign was active as of January 2026, during which victims were served a fake installer from 7zip[.]com, a highly convincing typo-squatted domain impersonating the official 7‑Zip distribution site (7-zip[.]org).

Initial access is typically achieved through social engineering and search‑engine abuse, including YouTube tutorial content that explicitly referenced the impersonated domain as the download source. Notably, several reports observed the installer delivered a modified but functional build of 7‑Zip (7zfm.exe) to reduce suspicion and preserve expected user behavior.

However, the installer also dropped additional payloads, such as Uphero.exe, hero.exe, and hero.dll, which are not part of the legitimate 7‑Zip software package. Once installed and executed, these payloads allow the attacker to establish persistence and configure the infected host as a proxy node under their control. This facilitates malicious activities such as traffic relaying, anonymizing infrastructure, and the delivery of secondary payloads [1] [2].

Overall, this attack illustrates a proxyware-style attack that abuses implicit trust in widely deployed third‑party tools while exploiting unconventional delivery vectors such as instructional media. By closely imitating legitimate software behavior and branding, the threat actors significantly reduced user suspicion and increased the likelihood of widespread, undetected compromise.

Threat overview

Darktrace observed multiple customers affected by the malicious 7‑Zip installer between January 12 and January 22, impacting organizations across the Americas (AMS), Asia‑Pacific & Japan (APJ), and Europe, the Middle East, and Africa (EMEA) regions. The activity targeted customers across various sectors, including Human health and social work activities, Manufacturing, Education, and Information and communication.

The following use case highlights a device on one customer network making external connections associated with malicious 7-Zip update activity observed between  January 7 and January 18, 2026.  This behavior included connectivity to the malicious domain 7zip[.]com, followed by command-and control (C2) activity involving "smshero"-themed domains, as well as outbound proxy connections over ports 1000 and 1002.  

Initial Connectivity to 'update[.]7zip[.]com':

Initial Beaconing to Young Endpoint alert behavior, involving the known tunnel/proxy endpoint ‘79.127.221[.]47’.
Figure 1: Initial Beaconing to Young Endpoint alert behavior, involving the known tunnel/proxy endpoint ‘79.127.221[.]47’.

Starting on January 7, Darktrace / NETWORK detected the device making repeated beaconing connections to the endpoint 79.127.221[.]47 over the destination port 1000. The use of this port aligns with open-source intelligence (OSINT) reporting that hero[.]exe establishes outbound proxy connections via non-standard ports such as 1000 and 1002 [1].

Darktrace observed TLS beaconing alerts to the known trojanized installer, update[.]7zip[.]com · 98.96.229[.]19, over port 443 on January 7th.
Figure 2: Darktrace observed TLS beaconing alerts to the known trojanized installer, update[.]7zip[.]com · 98.96.229[.]19, over port 443 on January 7th.

Later the same day, the device initiated TLS beaconing to the endpoint update.7zip[.]com. This is more than likely a common source of compromise, where victims unknowingly installed a modified build of the tool alongside additional malicious components. The campaign then progressed into the next attack phase, marked by established connectivity to various C2 domains.

Beaconing Activity to "smshero"-themed domains

Darktrace subsequently observed the same infected device connecting to various C2 domains used to retrieve configuration data. As such, these external hostnames were themed around the string “smshero”, for example ‘smshero[.]co’.

On January 8th, Darktrace observed SSL beaconing to a rare destination which was attributed to a known ‘config/control domain’, nova[.]smshero[.]ai.
Figure 3: On January 8th, Darktrace observed SSL beaconing to a rare destination which was attributed to a known ‘config/control domain’, nova[.]smshero[.]ai.

The following day, on January 8, the device exhibited its first connectivity to a "smshero"-themed endpoint, which has since been identified as being associated with rotating C2 servers [1] [3]. Similar beaconing activity continued over the following days, with Darktrace identifying C2 connectivity to update[.]7zip[.]com over port 443, alongside additional connections to “smshero”‑themed endpoints such as zest.hero-sms[.]ai, flux.smshero[.]cc, and glide.smshero[.]cc between January 9 and January 15.

Darktrace later observed continued beaconing alerts over a 4-day interval to additional rare destinations attributed to a known ‘config/control domain’, zest[.]hero-sms[.]ai & glide[.]smshero[.]cc.
Figure 4: Darktrace later observed continued beaconing alerts over a 4-day interval to additional rare destinations attributed to a known ‘config/control domain’, zest[.]hero-sms[.]ai & glide[.]smshero[.]cc.

Proxied connectivity over destination ports

The primary objective of this campaign is believed to be proxyware, whereby third-party traffic is routed through victim devices to potentially obfuscate malicious activity. Devices were also observed communicating with rare external IPs hosted on Cloudflare and DataCamp Limited ASNs, establishing outbound proxy connections over the non-standard ports 1000 and 1002 [1].

OSINT sources also indicate that connections over these ports leveraged an XOR-encoded protocol (key 0x70) designed to obscure control messages. While the end goal of the campaign remains unclear, residential proxy networks can be abused to evade security rules and facilitate further unauthorized activities, including phishing and malware distribution [1][3].

Specifically, on January 8, Darktrace observed the device engaging in low-and-slow data exfiltration to the IP 79.127.221[.]47, which had first been observed the previous day, over port 1000. Proxyware typically installs an agent that routes third‑party traffic through an end-user’s device, effectively  turning it into a residential proxy exit node. This activity likely represents the system actively communicating outbound data to an entity that controls its behavior.

Figure 5: Darktrace later observed a ‘Low and Slow Exfiltration to IP’ alert, involving the known tunnel/proxy endpoint ‘79.127.221[.]47’.

Similar activity continued between January 10 and January 18, with Darktrace detecting threat actors attempting to exfiltrate significant volumes of data to 79.127.221[.]47 over destination port 1000.

Throughout the course of this incident, Darktrace’s Cyber AI Analyst launched several autonomous investigations, analyzing each anomalous event and ultimately painting a detailed picture of the attack timeline. These investigations correlated multiple incidents based on Darktrace detections observed between January 7 and January 19. Cyber AI Analyst identified anomalous variables such as repeated connections to unusual endpoints involving data uploads and downloads, with particular emphasis on HTTP and SSL connectivity.

Darktrace AI Analyst Coverage, showcasing multiple incident events that occurred on January 7th & 8th, highlighting associated malicious 7-zip behaviors.
Figure 6: Darktrace AI Analyst Coverage, showcasing multiple incident events that occurred on January 7th & 8th, highlighting associated malicious 7-zip behaviors.
Darktrace AI Analyst Endpoint Details from the given ‘Unusual Repeated Connections’ Incident Event, including the known tunnel/proxy endpoint.
Figure 7: Darktrace AI Analyst Endpoint Details from the given ‘Unusual Repeated Connections’ Incident Event, including the known tunnel/proxy endpoint.
 Darktrace AI Analyst Coverage, showcasing additional incident events that occurred on January 12th through 18th, highlighting malicious 7-zip behaviors and SSL connectivity.
Figure 8: Darktrace AI Analyst Coverage, showcasing additional incident events that occurred on January 12th through 18th, highlighting malicious 7-zip behaviors and SSL connectivity.

Darktrace’s Autonomous Response

At several stages throughout the attack, Darktrace implemented Autonomous Response actions to help contain the suspicious activity as soon as it was identified, providing the customer’s security team with additional time to investigate and remediate. Between January 7 and January 18, Darktrace blocked a wide range of malicious activity, including beaconing connections to unusual endpoints, small data exfiltration attempts, and larger egress efforts, ultimately preventing the attacker from progressing through multiple stages of the attack or achieving their objectives.

Darktrace Autonomous Response Action Coverage showcasing connection block connection events including various endpoints that occurred on January 7th.
Figure 9: Darktrace Autonomous Response Action Coverage showcasing connection block connection events including various endpoints that occurred on January 7th.
Darktrace Antigena (Autonomous Response) Model Alert Coverage, showcasing a Antigena Suspicious Activity Block alert occurred on January 10th as a result of the Low and Slow Exfiltration to IP model alert.
Figure 10: Darktrace Antigena (Autonomous Response) Model Alert Coverage, showcasing a Antigena Suspicious Activity Block alert occurred on January 10th as a result of the Low and Slow Exfiltration to IP model alert.
Figure 11: Additional Darktrace Antigena (Autonomous Response) Model Alert Coverage, showcasing a Antigena Large Data Volume Outbound Block alert occurred on January 18th as a result of the Uncommon 1 GiB Outbound model alert.

Conclusion

The malicious 7‑Zip installer underscores how attackers continue to weaponize trust in widely used, legitimate software to gain initial access while evading user suspicion. By exploiting familiar and commonly installed services, this type of attack demonstrates that even routine actions, such as installing compression software, can become high‑risk events when defenses or user awareness are insufficient.

This campaign further emphasizes the urgent need for strict software validation and continuous network monitoring. Modern threats no longer rely solely on obscure tools or overtly malicious behavior. Instead, they increasingly blend seamlessly into everyday operations, making detection more challenging.

In this case, Darktrace / NETWORK was able to identify the anomalous activity and Autonomous Response actions in a timely manner, enabling the customer to be quickly notified and providing crucial additional time to investigate further.

In summary, the abuse of a trojanized 7‑Zip installer highlights a concerning shift in modern threat tactics, where trusted and widely deployed tools can serve as primary delivery mechanisms for system compromise. This reality reinforces that proactive detection, continuous monitoring, and strong security awareness are not optional but essential.

Credit to Justin Torres, Senior Cyber Analyst, David Moreira da Silva, Cyber Analyst, Emma Foulger, Global Threat Research Operations Lead.

Edited by Ryan Traill (Content Manager)

Appendices

References

1. https://www.malwarebytes.com/blog/threat-intel/2026/02/fake-7-zip-downloads-are-turning-home-pcs-into-proxy-nodes

2. https://www.tomshardware.com/tech-industry/cyber-security/unofficial-7-zip-com-website-served-up-malware-for-10-days-files-turned-pcs-into-a-proxy-botnet

3. https://blog.lukeacha.com/2026/01/beware-of-fake-7zip-installer-upstage.html

4. https://www.bleepingcomputer.com/news/security/malicious-7-zip-site-distributes-installer-laced-with-proxy-tool/

5. https://customerportal.darktrace.com/guides/antigena-network-model-actions

Darktrace Model Detections

·      Anomalous Connection / Data Sent to Rare Domain

·      Anomalous Connection / Low and Slow Exfiltration to IP

·      Anomalous Connection / Multiple Failed Connections to Rare Endpoint

·      Anomalous Connection / Uncommon 1 GiB Outbound

·      Anomalous Server Activity / Rare External from Server

·      Compromise / Agent Beacon (Long Period)

·      Compromise / Beacon for 4 Days

·      Compromise / Beacon to Young Endpoint

·      Compromise / Beaconing Activity To External Rare

·      Compromise / High Volume of Connections with Beacon Score

·      Compromise / Large Number of Suspicious Failed Connections

·      Compromise / Large Number of Suspicious Successful Connections

·      Compromise / Repeating Connections Over 4 Days

·      Compromise / SSL Beaconing to Rare Destination

·      Compromise / Suspicious TLS Beaconing To Rare External

·      Device / Large Number of Model Alerts

·      Unusual Activity / Unusual External Activity

Cyber AI Analyst Coverage

·      Unusual Repeated Connections

·      Unusual Repeated Connections to Multiple Endpoints

·      Possible HTTP Command and Control

·      Possible HTTP Command and Control to Multiple Endpoints

·      Suspicious Remote Service Control Activity

·      Possible SSL Command and Control to Multiple Endpoints

Indicators of Compromise

IoC - Type - Description + Confidence

·      7zip[.]com – Hostname – C2 Endpoint

·      flux[.]smshero[.]co - Hostname - C2 Endpoint

·      neo[.]herosms[.]co - Hostname - C2 Endpoint

·      nova[.]smshero[.]ai - Hostname - C2 Endpoint

·      zest[.]hero-sms[.]ai -  Hostname - C2 Endpoint

·      soc[.]hero-sms[.]co - Hostname - C2 Endpoint

·      pulse[.]herosms[.]cc - Hostname - C2 Endpoint

·      glide[.]smshero[.]cc - Hostname - C2 Endpoint

·      prime[.]herosms[.]vip - Hostname - C2 Endpoint

·      172.96.115[.]226 - IP Address - C2 Endpoint

·      79.127.221[.]47:1002 – IP Address/Port - Proxy Endpoint

·      84.17.37[.]1:1002 - IP Address/Port - Proxy Endpoint

MITRE ATT&CK Mapping

Technique Name - Tactic - ID - Sub-Technique of

·      Exfiltration Over C2 Channel - EXFILTRATION - T1041

·      Scheduled Transfer - EXFILTRATION - T1029

·      Automated Exfiltration - EXFILTRATION - T1020

·      Data Transfer Size Limits - EXFILTRATION - T1030

·      External Proxy - COMMAND AND CONTROL - T1090.002 - T1090

·      Non-Application Layer Protocol - COMMAND AND CONTROL - T1095

·      Non-Standard Port - COMMAND AND CONTROL - T1571

·      Exfiltration to Cloud Storage - EXFILTRATION - T1567.002 - T1567

·      Exploit Public-Facing Application - INITIAL ACCESS - T1190

·      Web Protocols - COMMAND AND CONTROL - T1071.001 - T1071

·      Application Layer Protocol - COMMAND AND CONTROL - T1071

·      Man in the Browser - COLLECTION - T1185

·      Browser Extensions - PERSISTENCE - T1176

·      Encrypted Channel - COMMAND AND CONTROL - T1573

·      Fallback Channels - COMMAND AND CONTROL - T1008

·      Multi-Stage Channels - COMMAND AND CONTROL - T1104

·      Supply Chain Compromise - INITIAL ACCESS ICS - T0862

·      Commonly Used Port - COMMAND AND CONTROL ICS - T0885

Inside the SOC
Darktrace cyber analysts are world-class experts in threat intelligence, threat hunting and incident response, and provide 24/7 SOC support to thousands of Darktrace customers around the globe. Inside the SOC is exclusively authored by these experts, providing analysis of cyber incidents and threat trends, based on real-world experience in the field.
Written by
Justin Torres
Cyber Analyst

More in this series

No items found.

Blog

/

Network

/

May 22, 2026

Darktrace named a Leader in the 2026 Gartner® Magic Quadrant™ for Network Detection and Response (NDR) For the Second Consecutive Year

garnter ndr magic quadrantDefault blog imageDefault blog image

Continued recognition in NDR  

Darktrace has been recognized as a Leader in the 2026 Gartner® Magic Quadrant™ for Network Detection and Response (NDR), marking the second consecutive year in the Leaders quadrant.

We believe this consistency reflects sustained ability to execute, adapt, and deliver outcomes as the market evolves.

While we are immensely proud to be recognized by industry analysts as a Leader in NDR, that's just part of the story. Darktrace was also Named the Only 2025 Gartner® Peer Insights™ Customers’ Choice for Network Detection and Response based on direct customer feedback and real-world experience.

We believe the combination of these two signals is important. One reflects how the market is evaluated. The other reflects how technology performs in practice.

Why Darktrace continues to be recognized as a leader

We believe our position as a Leader for the second consecutive year reflects a combination of our sustained ability to execute in NDR, continued AI innovation, and proven delivery of security outcomes for customers and partners worldwide.

We also feel that our leadership in the NDR market is a testament to our unique and multi-layered AI approach, for which we were recognized as No.7 on Fast Company’s Most Innovative AI Companies of 2026 list, plus one of the hottest AI cybersecurity companies in CRN's AI 100.

Adapting to complex, real-world environments

Organizations are no longer protecting a single network perimeter. They are securing a mix of users, devices, applications, and data that move across hybrid environments.

Darktrace has focused on maintaining visibility and detection across these conditions, allowing security teams to understand activity as it scales.

Supporting organizations globally, not just technically

Security outcomes are shaped as much by deployment and support as they are by detection capability.

Darktrace continues to invest in regional presence across 29 countries around the world, helping organizations operationalize NDR in ways that align with local requirements, internal processes, and team structures.

Continuing to push AI beyond detection

AI in cybersecurity is often positioned as a way to improve detection accuracy. But the more important shift is how AI can influence decision-making and response.

Darktrace continues to develop models that learn from both live environments and historical incident data, combining real-time behavioral analysis with insights derived from prior attack patterns.

Using technologies such as the Incident Graph and DIGEST (Darktrace Incident Graph Evaluation for Security Threats), activity is not analyzed in isolation. Instead, relationships between users, devices, connections, and events are mapped over time, allowing the system to reconstruct how an incident is unfolding and how similar incidents have progressed in the past.

By evaluating these patterns, Darktrace can assess the likelihood that an incident will escalate, prioritizing the activity that poses the greatest risk and surfacing the most relevant context for investigation.

This shifts security operations from simply identifying anomalies to understanding their trajectory, helping teams anticipate potential impact and respond earlier with greater precision.

Why NDR is shifting from reactive detection to proactive, AI-driven security

Traditional approaches to NDR have been built around reactively identifying threats once they become clearly visible. That model is increasingly difficult to rely on.

Attackers are no longer operating in ways that stand out. They use valid credentials, trusted tools, and low-and-slow techniques that blend into everyday activity. By the time something looks obviously malicious, the impact is often already underway.

This is the core limitation of reactive detection. It depends on recognizing something that already looks like a threat.

As a result, many of the most consequential incidents today fall into a gap.

Insider activity, compromised credentials, and novel attacks rarely trigger traditional alerts because they do not follow known patterns. On the surface, they often appear legitimate, making them difficult to distinguish from normal behavior without deeper context.

This is why we believe this Gartner recognition reflects a broader shift in NDR toward autonomous, proactive and pre‑emptive security operations.

By understanding normal behavior within an environment, it is possible to identify subtle deviations rather than waiting for confirmation of threats as they are taking place.

Darktrace’s Self-Learning AI is designed for behavioral understanding. By continuously learning each organization’s normal patterns, it can detect deviations in real time, enabling a proactive and pre-emptive model of NDR where security teams can respond to early signs of risk as they emerge, reducing the window in which attacks can develop.

In multiple cases, this behavioral approach has led to early threat detection where Darktrace identified completely unknown threats, including pre-CVE zero-day activity. By detecting subtle behavioral changes before vulnerabilities were publicly disclosed or widely understood, organizations can mitigate threats before they do damage.

This shift is subtle but important. Modern NDR solutions must shift from a system that explains what happened to one that helps prevent threats from developing in the first place, and Darktrace is proud to be at the forefront of this shift - helping organizations build and maintain a state of proactive network resilience.

Continuing to innovate at the forefront of NDR

In our view, recognition as a Leader reflects where the market is today. Continuing to innovate defines what comes next.

As businesses evolve, new technologies like AI tools and agents introduce new security risks and challenges; security teams need more than simple detection. They need a complete understanding of risk as it develops, the ability to investigate it in context, and to contain threats at machine speed.  

Darktrace / NETWORK is built to deliver across that full spectrum. Its Self-Learning AI continuously adapts to each organization’s environment, identifying subtle behavioral changes that signal emerging threats. Integrated investigation and autonomous response reduce the time between detection and action, allowing teams to move with greater speed and confidence.

This combination enables organizations to detect and contain known, unknown, and insider threats as they develop, while also strengthening resilience over time.

As a two-time Leader in the Gartner® Magic Quadrant™ for NDR and the only 2025 Gartner® Peer Insights™ Customers’ Choice, we feel Darktrace continues to evolve its platform to meet the demands of modern environments, delivering a more complete and adaptive approach to network security.

[related-resource]

Disclaimer: The 2026 Gartner® Magic Quadrant™ for Network Detection and Response (NDR) ,The 2026 Gartner® Magic Quadrant™ for Network Detection and Response (NDR), Thomas Lintemuth, Charanpal Bhogal, Nahim Fazal, 18 May 2026.

Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner’s research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.

GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved. Magic Quadrant is a registered trademark of Gartner, Inc. and/or its affiliates and is used herein with permission. All rights reserved.

Continue reading
About the author
Mikey Anderson
Product Marketing Manager, Network Detection & Response

Blog

/

AI

/

May 21, 2026

Prompt Security in Enterprise AI: Strengths, Weaknesses, and Common Approaches

prompt securityDefault blog imageDefault blog image

How enterprise AI Agents are changing the risk landscape  

Generative AI Agents are changing the way work gets done inside enterprises, and subsequently how security risks may emerge. Organizations have quickly realized that providing these agents with wider access to tooling, internal information, and granting permissions for the agent to perform autonomous actions can greatly increase the efficiency of employee workflows.

Early deployments of Generative AI systems led many organizations to scope individual components as self-contained applications: a chat interface, a model, and a prompt, with guardrails placed at the boundary. Research from Gartner has shown that while the volume and scope of Agentic AI deployments in enterprise environments is rapidly accelerating, many of the mechanisms required to manage risk, trust, and cost are still maturing.

The issue now resides on whether an agent can be influenced, misdirected, or manipulated in ways that leads to unsafe behavior across a broader system.

Why prompt security matters in enterprise AI

Prompt security matters in enterprise AI because prompts are the primary way users and systems interact with Agentic AI models, making them one of the earliest and most visible indicators of how these systems are being used and where risk may emerge.

For security teams, prompt monitoring is a logical starting point for understanding enterprise AI usage, providing insight into what types of questions are being asked and tasks are being given to AI Agents, how these systems are being guided, and whether interactions align with expected behavior. Complete prompt security takes this one step further, filtering out or blocking sensitive or dangerous content to prevent risks like prompt injection and data leakage.

However, visibility only at the prompt layer can create a false sense of security. Prompts show what was asked, but not always why it was asked, or what downstream actions were triggered by the agent across connected systems, data sources, or applications.

What prompt security reveals  

The primary function of prompt security is to minimize risks associated with generative and agentic AI use, but monitoring and analysis of prompts can also grant insight into use cases for particular agents and model. With comprehensive prompt security, security teams should be able to answer the following questions for each prompt:

  • What task was the user attempting to complete?
  • What data was included in the request, and was any of the data high-risk or confidential?
  • Was the interaction high-risk, potentially malicious, or in violation of company policy?
  • Was the prompt anomalous (in comparison to previous prompts sent to the agent / model)?

Improving visibility at this layer is a necessary first step, allowing organizations to establish a baseline for how AI systems are being used and where potential risks may exist.  

Prompt security alone does not provide a complete view of risk. Further data is needed to understand how the prompt is interpreted, how context is applied, what autonomous actions the agent takes (if any), or what downstream systems are affected. Understanding the outcome of a query is just as important for complete prompt security as understanding the input prompt itself – for example, a perfectly normal, low-risk prompt may inadvertently result in an agent taking a high-risk action.

Comprehensive AI security systems like Darktrace / SECURE AI can monitor and analyze both the prompt submitted to a Generative AI system, as well as the responses and chain-of-thought of the system, providing greater insight into the behavior of the system. Darktrace / SECURE AI builds on the core Darktrace methodology, learning the expected behaviors of your organization and identifying deviations from the expected pattern of life.

How organizations address prompt security today

As prompt-level visibility has become a focus, a range of approaches have emerged to make this activity more observable and controllable. Various monitoring and logging tools aim to capture prompt inputs to be analyzed after the fact.  

Input validation and filtering systems attempt to intervene earlier, inspecting prompts before they reach the model. These controls look for known jailbreak patterns, language indicative of adversarial attacks, or ambiguous instructions which could push the system off course.

Importantly, for a prompt security solution to be accurate and effective, prompts must be continually observed and governed, rather than treated as a point-in-time snapshot.  

Where prompt security breaks down in real environments

In more complex environments, especially those involving multiple agents or extensive tool use, AI security becomes harder to define and control.

Agent-to-Agent communications can be harder to monitor and trace as these happen without direct user interaction. Communication between agents can create routes for potential context leakage between agents, unintentional privilege escalation, or even data leakage from a higher privileged agent to a lower privileged one.

Risk is shaped not just by what is asked, but by the conditions in which that prompt operates and the actions an agent takes. Controls at the orchestration layer are starting to reflect this reality. Techniques such as context isolation, scoped memory, and role-based boundaries aim to limit how far a prompt’s influence can extend.  

Furthermore, Shadow AI usage can be difficult to monitor. AI systems that are deployed outside of formal governance structures and Generative AI systems hosted on unknown endpoints can fly under the radar and can go unseen by monitoring tools, leaving a critical opening where adversarial prompts may go undetected. Darktrace / SECURE AI features comprehensive detection of Shadow AI usage, helping organizations identify potential risk areas.

How prompt security fits in a broader AI risk model

Prompt security is an important starting point, but it is not a complete security strategy. As AI systems become more integrated into enterprise environments, the risks extend to what resources the system can access, how it interprets context, and what actions it is allowed to take across connected tools and workflows.

This creates a gap between visibility and control. Prompt security alone allows security teams to observe prompt activity but falls short of creating a clear understanding of how that activity translates into real-world impact across the organization.

Closing that gap requires a broader approach, one that connects signals across human and AI agent identities, SaaS, cloud, and endpoint environments. It means understanding not just how an AI system is being used, but how that usage interacts with the rest of the digital estate.

Prompt security, in that sense, is less of a standalone solution and more of an entry point into a larger problem: securing AI across the enterprise as a whole.

Explore how Darktrace / SECURE AI brings prompt security to enterprises

Darktrace brings more than a decade of AI expertise, built on an enterprise‑wide platform designed to operate in and understand the behaviors of the complex, ambiguous environments where today’s AI now lives. With Darktrace / SECURE AI, enterprises can safely adopt, manage, monitor, and build AI within their business.  

Learn about Darktrace / SECURE AI here.

Sign up today to stay informed about innovations across securing AI.

[related-resource]

Continue reading
About the author
Jamie Bali
Technical Author (AI) Developer
Your data. Our AI.
Elevate your network security with Darktrace AI