Blog
/
/
February 9, 2021

Antigena Email Version 5: The Future of Email Protection

Version 5 of Antigena Email enhances security operations with AI-powered threat detection and intuitive reporting for busy security teams.
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
Dan Fein
VP, Product
Default blog imageDefault blog imageDefault blog imageDefault blog imageDefault blog imageDefault blog image
09
Feb 2021

Darktrace Version 5 signals a new chapter in AI-powered cyber security, offering a series of innovations across the entire Immune System platform – including AI augmentation and extended coverage across remote environments. This update also includes one-click integrations, on-demand automated investigations, and – the subject of this blog post – critical upgrades to Antigena Email, the world’s first autonomous email security technology.

Antigena Email uses a self-learning approach to stop every type of email threat, without relying on pre-existing lists or reputation checks. The technology autonomously interrogates every email in the context of its evolving understanding of ‘normal’ for the recipient, group, and organization as a whole. The features in Version 5 present several unique benefits to the user, not least in the various ways in which they can save time.

The self-learning AI technology provides a solution free from configuration, policy setting, and ongoing maintenance. The system’s accuracy results in negligible false positives, meaning security teams no longer need to release legitimate emails that legacy security tools have held back.

Furthermore, human security teams are augmented by Narrative – a new feature that automatically generates natural language reports on every email security incident. By surfacing a summary of what happened and why Antigena Email took the actions it did, Version 5 drastically reduces ‘time to meaning’ for overstretched human security teams.

Time to resolve a phishing attack

Email attacks are becoming increasingly targeted, and just one successful attack can give hackers the keys to an organization’s digital kingdom. Investigating the cause of a breach, cleaning up infected devices, and manually compiling incident reports can quickly drain a company’s resources.

Gateway tools tend to be time-consuming for security professionals, who must research malicious emails that were let through and tweak settings to stop them in the future, as well as release ‘false positive’ legitimate business emails that have been stopped for no good reason. Under such constraints, it is no wonder that phishing emails are reaching the inbox with alarming frequency – leading to wide-scale attacks.

While many traditional security tools put immense strain on human analysts, Antigena Email almost entirely removes the human from the equation. The self-learning technology accurately determines malicious from benign by taking a fundamentally different approach to email security. Rather than asking ‘is this email bad’ – Antigena Email uniquely sets to find out: ‘does this email belong’, in the context of ‘normal’ for the sender, the recipient, and the wider organization. It is this contextual understanding of the wider ‘patterns of life’ that enables the technology to catch sophisticated threats on the first encounter.

Time to find and release emails

Security teams too often spend their days ground down by repetitive tasks. For those who rely on legacy tools which present crude information and stop only the most basic threats, important trends are not found unless manually uncovered, and human experts are kept in the weeds.

With Antigena Email, this has now changed. Customers are now able to focus on gaining a holistic understanding of their organization. Such understanding is only possible when teams are not bogged down in details or trapped by an obscure user interface, tweaking complex settings which could inadvertently cause more harm than good.

The technology generates a bespoke dashboard for security teams, accounting for all specific preferences and interests. For example, organizations interested primarily in supply chain attacks on the C-suite can set Antigena Email to surface and chart anomalous emails tagged by Antigena Email as ‘Out of Character’, where specifically the recipient was C-suite.

Figure 1: With Antigena Email Version 5, there is no need to log in and no action to be taken. When users do log in, they are presented with high-level metrics of the email threats facing their organization.

In this way, IT teams can set the system once to exactly what interests them, and subsequently forget about it until they decide to log in and glance over key figures. When logging in, it is no longer to chase a specific email, and there is nothing to action – Antigena Email has already done it. Instead, IT teams can view the broad picture and use the information available to influence security decisions. They can now ask and fully understand which users are most exposed and why an organization is so at risk.

Time to understand what happened

Security professionals just need the answer. When looking at an email, no one should have to unpack and make sense of raw data. Instead, users should be presented with a recap summary – a Narrative – which is digestible in seconds and which even the most junior team members can easily grasp.

Antigena Email takes each complex case and words it in such a way that even a non-technical employee can understand. It uses advanced machine learning to present key information in plain English, allowing end users to perceive the situation at a glance.

Figure 2: An example of Antigena Email’s Narrative summary on the right hand side of the screen

Narrative tells the stories of what happened and why, and how aggressively an email was actioned. What was the sender’s intention? Were they trying to solicit the recipient into a bank transaction? Whatever the circumstances, if an email does not belong, that is the end of the story. There are no ongoing chapters, there is no fallout. Antigena Email neutralizes the email and ends the story before the threat has had the chance to develop.

And if a person wishes to dive deeper, Narrative provides one-click jumping off points that expose the underlying data (see the red text in the image above). But this is a choice. It is no longer business critical to scroll through emails and uncover information manually to stop future threats. As Antigena Email is proactive, the human no longer has to be.

A new era of email security

Antigena Email takes care of all the daily repetitive tasks – stopping the bad, allowing the good – taking the least aggressive action to neutralize any given threat. As a result, security teams are no longer forced to spend their days determining which emails are malicious or dealing with complaints from users who have had legitimate emails blocked.

Now that human experts no longer have to worry about sifting through emails themselves, they can focus on what matters. Antigena Email gives time to security teams to define their email environment, pinpoint the biggest risks, and identify general business trends.

Find out more about Darktrace Version 5

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
Dan Fein
VP, Product

More in this series

No items found.

Blog

/

Cloud

/

January 15, 2026

React2Shell Reflections: Cloud Insights, Finance Sector Impacts, and How Threat Actors Moved So Quickly

React2Shell Default blog imageDefault blog image

Introduction

Last month’s disclosure of CVE 2025-55812, known as React2Shell, provided a reminder of how quickly modern threat actors can operationalize newly disclosed vulnerabilities, particularly in cloud-hosted environments.

The vulnerability was discovered on December 3, 2025, with a patch made available on the same day. Within 30 hours of the patch, a publicly available proof-of-concept emerged that could be used to exploit any vulnerable server. This short timeline meant many systems remained unpatched when attackers began actively exploiting the vulnerability.  

Darktrace researchers rapidly deployed a new honeypot to monitor exploitation of CVE 2025-55812 in the wild.

Within two minutes of deployment, Darktrace observed opportunistic attackers exploiting this unauthenticated remote code execution flaw in React Server Components, leveraging a single crafted request to gain control of exposed Next.js servers. Exploitation quickly progressed from reconnaissance to scripted payload delivery, HTTP beaconing, and cryptomining, underscoring how automation and pre‑positioned infrastructure by threat actors now compress the window between disclosure and active exploitation to mere hours.

For cloud‑native organizations, particularly those in the financial sector, where Darktrace observed the greatest impact, React2Shell highlights the growing disconnect between patch availability and attacker timelines, increasing the likelihood that even short delays in remediation can result in real‑world compromise.

Cloud insights

In contrast to traditional enterprise networks built around layered controls, cloud architectures are often intentionally internet-accessible by default. When vulnerabilities emerge in common application frameworks such as React and Next.js, attackers face minimal friction.  No phishing campaign, no credential theft, and no lateral movement are required; only an exposed service and exploitable condition.

The activity Darktrace observed during the React2shell intrusions reflects techniques that are familiar yet highly effective in cloud-based attacks. Attackers quickly pivot from an exposed internet-facing application to abusing the underlying cloud infrastructure, using automated exploitation to deploy secondary payloads at scale and ultimately act on their objectives, whether monetizing access through cryptomining or to burying themselves deeper in the environment for sustained persistence.

Cloud Case Study

In one incident, opportunistic attackers rapidly exploited an internet-facing Azure virtual machine (VM) running a Next.js application, abusing the React/next.js vulnerability to gain remote command execution within hours of the service becoming exposed. The compromise resulted in the staged deployment of a Go-based remote access trojan (RAT), followed by a series of cryptomining payloads such as XMrig.

Initial Access

Initial access appears to have originated from abused virtual private network (VPN) infrastructure, with the source IP (146.70.192[.]180) later identified as being associated with Surfshark

The IP address above is associated with VPN abuse leveraged for initial exploitation via Surfshark infrastructure.
Figure 1: The IP address above is associated with VPN abuse leveraged for initial exploitation via Surfshark infrastructure.

The use of commercial VPN exit nodes reflects a wider trend of opportunistic attackers leveraging low‑cost infrastructure to gain rapid, anonymous access.

Parent process telemetry later confirmed execution originated from the Next.js server, strongly indicating application-layer compromise rather than SSH brute force, misused credentials, or management-plane abuse.

Payload execution

Shortly after successful exploitation, Darktrace identified a suspicious file and subsequent execution. One of the first payloads retrieved was a binary masquerading as “vim”, a naming convention commonly used to evade casual inspection in Linux environments. This directly ties the payload execution to the compromised Next.js application process, reinforcing the hypothesis of exploit-driven access.

Command-and-Control (C2)

Network flow logs revealed outbound connections back to the same external IP involved in the inbound activity. From a defensive perspective, this pattern is significant as web servers typically receive inbound requests, and any persistent outbound callbacks — especially to the same IP — indicate likely post-exploitation control. In this case, a C2 detection model alert was raised approximately 90 minutes after the first indicators, reflecting the time required for sufficient behavioral evidence to confirm beaconing rather than benign application traffic.

Cryptominers deployment and re-exploitation

Following successful command execution within the compromised Next.js workload, the attackers rapidly transitioned to monetization by deploying cryptomining payloads. Microsoft Defender observed a shell command designed to fetch and execute a binary named “x” via either curl or wget, ensuring successful delivery regardless of which tooling was availability on the Azure VM.

The binary was written to /home/wasiluser/dashboard/x and subsequently executed, with open-source intelligence (OSINT) enrichment strongly suggesting it was a cryptominer consistent with XMRig‑style tooling. Later the same day, additional activity revealed the host downloading a static XMRig binary directly from GitHub and placing it in a hidden cache directory (/home/wasiluser/.cache/.sys/).

The use of trusted infrastructure and legitimate open‑source tooling indicates an opportunistic approach focused on reliability and speed. The repeated deployment of cryptominers strongly suggests re‑exploitation of the same vulnerable web application rather than reliance on traditional persistence mechanisms. This behavior is characteristic of cloud‑focused attacks, where publicly exposed workloads can be repeatedly compromised at scale more easily.

Financial sector spotlight

During the mass exploitation of React2Shell, Darktrace observed targeting by likely North Korean affiliated actors focused on financial organizations in the United Kingdom, Sweden, Spain, Portugal, Nigeria, Kenya, Qatar, and Chile.

The targeting of the financial sector is not unexpected, but the emergence of new Democratic People’s Republic of Korea (DPRK) tooling, including a Beavertail variant and EtherRat, a previously undocumented Linux implant, highlights the need for updated rules and signatures for organizations that rely on them.

EtherRAT uses Ethereum smart contracts for C2 resolution, polling every 500 milliseconds and employing five persistence mechanisms. It downloads its own Node.js runtime from nodejs[.]org and queries nine Ethereum RPC endpoints in parallel, selecting the majority response to determine its C2 URL. EtherRAT also overlaps with the Contagious Interview campaign, which has targeted blockchain developers since early 2025.

Read more finance‑sector insights in Darktrace’s white paper, The State of Cyber Security in the Finance Sector.

Threat actor behavior and speed

Darktrace’s honeypot was exploited just two minutes after coming online, demonstrating how automated scanning, pre-positioned infrastructure and staging, and C2 infrastructure traced back to “bulletproof” hosting reflects a mature, well‑resourced operational chain.

For financial organizations, particularly those operating cloud‑native platforms, digital asset services, or internet‑facing APIs, this activity demonstrates how rapidly geopolitical threat actors can weaponize newly disclosed vulnerabilities, turning short patching delays into strategic opportunities for long‑term access and financial gain. This underscores the need for a behavioral-anomaly-led security posture.

Credit to Nathaniel Jones (VP, Security & AI Strategy, Field CISO) and Mark Turner (Specialist Security Researcher)

Edited by Ryan Traill (Analyst Content Lead)

Appendices

Indicators of Compromise (IoCs)

146.70.192[.]180 – IP Address – Endpoint Associated with Surfshark

References

https://www.darktrace.com/resources/the-state-of-cybersecurity-in-the-finance-sector

Continue reading
About the author
Nathaniel Jones
VP, Security & AI Strategy, Field CISO

Blog

/

Cloud

/

January 13, 2026

Runtime Is Where Cloud Security Really Counts: The Importance of Detection, Forensics and Real-Time Architecture Awareness

runtime, cloud security, cnaapDefault blog imageDefault blog image

Introduction: Shifting focus from prevention to runtime

Cloud security has spent the last decade focused on prevention; tightening configurations, scanning for vulnerabilities, and enforcing best practices through Cloud Native Application Protection Platforms (CNAPP). These capabilities remain essential, but they are not where cloud attacks happen.

Attacks happen at runtime: the dynamic, ephemeral, constantly changing execution layer where applications run, permissions are granted, identities act, and workloads communicate. This is also the layer where defenders traditionally have the least visibility and the least time to respond.

Today’s threat landscape demands a fundamental shift. Reducing cloud risk now requires moving beyond static posture and CNAPP only approaches and embracing realtime behavioral detection across workloads and identities, paired with the ability to automatically preserve forensic evidence. Defenders need a continuous, real-time understanding of what “normal” looks like in their cloud environments, and AI capable of processing massive data streams to surface deviations that signal emerging attacker behavior.

Runtime: The layer where attacks happen

Runtime is the cloud in motion — containers starting and stopping, serverless functions being called, IAM roles being assumed, workloads auto scaling, and data flowing across hundreds of services. It’s also where attackers:

  • Weaponize stolen credentials
  • Escalate privileges
  • Pivot programmatically
  • Deploy malicious compute
  • Manipulate or exfiltrate data

The challenge is complex: runtime evidence is ephemeral. Containers vanish; critical process data disappears in seconds. By the time a human analyst begins investigating, the detail required to understand and respond to the alert, often is already gone. This volatility makes runtime the hardest layer to monitor, and the most important one to secure.

What Darktrace / CLOUD Brings to Runtime Defence

Darktrace / CLOUD is purpose-built for the cloud execution layer. It unifies the capabilities required to detect, contain, and understand attacks as they unfold, not hours or days later. Four elements define its value:

1. Behavioral, real-time detection

The platform learns normal activity across cloud services, identities, workloads, and data flows, then surfaces anomalies that signify real attacker behavior, even when no signature exists.

2. Automated forensic level artifact collection

The moment Darktrace detects a threat, it can automatically capture volatile forensic evidence; disk state, memory, logs, and process context, including from ephemeral resources. This preserves the truth of what happened before workloads terminate and evidence disappears.

3. AI-led investigation

Cyber AI Analyst assembles cloud behaviors into a coherent incident story, correlating identity activity, network flows, and Cloud workload behavior. Analysts no longer need to pivot across dashboards or reconstruct timelines manually.

4. Live architectural awareness

Darktrace continuously maps your cloud environment as it operates; including services, identities, connectivity, and data pathways. This real-time visibility makes anomalies clearer and investigations dramatically faster.

Together, these capabilities form a runtime-first security model.

Why CNAPP alone isn’t enough

CNAPP platforms excel at pre deployment checks all the way down to developer workstations, identifying misconfigurations, concerning permission combinations, vulnerable images, and risky infrastructure choices. But CNAPP’s breadth is also its limitation. CNAPP is about posture. Runtime defense is about behavior.

CNAPP tells you what could go wrong; runtime detection highlights what is going wrong right now.

It cannot preserve ephemeral evidence, correlate active behaviors across domains, or contain unfolding attacks with the precision and speed required during a real incident. Prevention remains essential, but prevention alone cannot stop an attacker who is already operating inside your cloud environment.

Real-world AWS Scenario: Why Runtime Monitoring Wins

A recent incident detected by Darktrace / CLOUD highlights how cloud compromises unfold, and why runtime visibility is non-negotiable. Each step below reflects detections that occur only when monitoring behavior in real time.

1. External Credential Use

Detection: Unusual external source for credential use: An attacker logs into a cloud account from a never-before-seen location, the earliest sign of account takeover.

2. AWS CLI Pivot

Detection: Unusual CLI activity: The attacker switches to programmatic access, issuing commands from a suspicious host to gain automation and stealth.

3. Credential Manipulation

Detection: Rare password reset: They reset or assign new passwords to establish persistence and bypass existing security controls.

4. Cloud Reconnaissance

Detection: Burst of resource discovery: The attacker enumerates buckets, roles, and services to map high value assets and plan next steps.

5. Privilege Escalation

Detection: Anomalous IAM update: Unauthorized policy updates or role changes grant the attacker elevated access or a backdoor.

6. Malicious Compute Deployment

Detection: Unusual EC2/Lambda/ECS creation: The attacker deploys compute resources for mining, lateral movement, or staging further tools.

7. Data Access or Tampering

Detection: Unusual S3 modifications: They alter S3 permissions or objects, often a prelude to data exfiltration or corruption.

Only some of these actions would appear in a posture scan, crucially after the fact.
Every one of these runtime detections is visible only through real-time behavioral monitoring while the attack is in progress.

The future of cloud security Is runtime-first

Cloud defense can no longer revolve solely around prevention. Modern attacks unfold in runtime, across a fast-changing mesh of workloads, services, and — critically — identities. To reduce risk, organizations must be able to detect, understand, and contain malicious activity as it happens, before ephemeral evidence disappears and before attacker's pivot across identity layers.

Darktrace / CLOUD delivers this shift by turning runtime, the most volatile and consequential layer in the cloud, into a fully defensible control point through unified visibility across behavior, workloads, and identities. It does this by providing:

  • Real-time behavior detection across workloads and identity activity
  • Autonomous response actions for rapid containment
  • Automated forensic level artifact preservation the moment events occur
  • AI-driven investigation that separates weak signals from true attacker patterns
  • Live cloud environment insight to understand context and impact instantly

Cloud security must evolve from securing what might go wrong to continuously understanding what is happening; in runtime, across identities, and at the speed attackers operate. Unifying runtime and identity visibility is how defenders regain the advantage.

[related-resource]

Continue reading
About the author
Adam Stevens
Senior Director of Product, Cloud | Darktrace
Your data. Our AI.
Elevate your network security with Darktrace AI