Blog
/
/
February 1, 2021

Explore AI Email Security Approaches with Darktrace

Default blog imageDefault blog imageDefault blog imageDefault blog imageDefault blog imageDefault blog image
01
Feb 2021
Stay informed on the latest AI approaches to email security. Explore Darktrace's comparisons to find the best solution for your cybersecurity needs!

Innovations in artificial intelligence (AI) have fundamentally changed the email security landscape in recent years, but it can often be hard to determine what makes one system different to the next. In reality, under that umbrella term there exists a significant distinction in approach which may determine whether the technology provides genuine protection or simply a perceived notion of defense.

One backward-looking approach involves feeding a machine thousands of emails that have already been deemed to be malicious, and training it to look for patterns in these emails in order to spot future attacks. The second approach uses an AI system to analyze the entirety of an organization’s real-world data, enabling it to establish a notion of what is ‘normal’ and then spot subtle deviations indicative of an attack.

In the below, we compare the relative merits of each approach, with special consideration to novel attacks that leverage the latest news headlines to bypass machine learning systems trained on data sets. Training a machine on previously identified ‘known bads’ is only advantageous in certain, specific contexts that don’t change over time: to recognize the intent behind an email, for example. However, an effective email security solution must also incorporate a self-learning approach that understands ‘normal’ in the context of an organization in order to identify unusual and anomalous emails and catch even the novel attacks.

Signatures – a backward-looking approach

Over the past few decades, cyber security technologies have looked to mitigate risk by preventing previously seen attacks from occurring again. In the early days, when the lifespan of a given strain of malware or the infrastructure of an attack was in the range of months and years, this method was satisfactory. But the approach inevitably results in playing catch-up with malicious actors: it always looks to the past to guide detection for the future. With decreasing lifetimes of attacks, where a domain could be used in a single email and never seen again, this historic-looking signature-based approach is now being widely replaced by more intelligent systems.

Training a machine on ‘bad’ emails

The first AI approach we often see in the wild involves harnessing an extremely large data set with thousands or millions of emails. Once these emails have come through, an AI is trained to look for common patterns in malicious emails. The system then updates its models, rules set, and blacklists based on that data.

This method certainly represents an improvement to traditional rules and signatures, but it does not escape the fact that it is still reactive, and unable to stop new attack infrastructure and new types of email attacks. It is simply automating that flawed, traditional approach – only instead of having a human update the rules and signatures, a machine is updating them instead.

Relying on this approach alone has one basic but critical flaw: it does not enable you to stop new types of attacks that it has never seen before. It accepts that there has to be a ‘patient zero’ – or first victim – in order to succeed.

The industry is beginning to acknowledge the challenges with this approach, and huge amounts of resources – both automated systems and security researchers – are being thrown into minimizing its limitations. This includes leveraging a technique called “data augmentation” that involves taking a malicious email that slipped through and generating many “training samples” using open-source text augmentation libraries to create “similar” emails – so that the machine learns not only the missed phish as ‘bad’, but several others like it – enabling it to detect future attacks that use similar wording, and fall into the same category.

But spending all this time and effort into trying to fix an unsolvable problem is like putting all your eggs in the wrong basket. Why try and fix a flawed system rather than change the game altogether? To spell out the limitations of this approach, let us look at a situation where the nature of the attack is entirely new.

The rise of ‘fearware’

When the global pandemic hit, and governments began enforcing travel bans and imposing stringent restrictions, there was undoubtedly a collective sense of fear and uncertainty. As explained previously in this blog, cyber-criminals were quick to capitalize on this, taking advantage of people’s desire for information to send out topical emails related to COVID-19 containing malware or credential-grabbing links.

These emails often spoofed the Centers for Disease Control and Prevention (CDC), or later on, as the economic impact of the pandemic began to take hold, the Small Business Administration (SBA). As the global situation shifted, so did attackers’ tactics. And in the process, over 130,000 new domains related to COVID-19 were purchased.

Let’s now consider how the above approach to email security might fare when faced with these new email attacks. The question becomes: how can you train a model to look out for emails containing ‘COVID-19’, when the term hasn’t even been invented yet?

And while COVID-19 is the most salient example of this, the same reasoning follows for every single novel and unexpected news cycle that attackers are leveraging in their phishing emails to evade tools using this approach – and attracting the recipient’s attention as a bonus. Moreover, if an email attack is truly targeted to your organization, it might contain bespoke and tailored news referring to a very specific thing that supervised machine learning systems could never be trained on.

This isn’t to say there’s not a time and a place in email security for looking at past attacks to set yourself up for the future. It just isn’t here.

Spotting intention

Darktrace uses this approach for one specific use which is future-proof and not prone to change over time, to analyze grammar and tone in an email in order to identify intention: asking questions like ‘does this look like an attempt at inducement? Is the sender trying to solicit some sensitive information? Is this extortion?’ By training a system on an extremely large data set collected over a period of time, you can start to understand what, for instance, inducement looks like. This then enables you to easily spot future scenarios of inducement based on a common set of characteristics.

Training a system in this way works because, unlike news cycles and the topics of phishing emails, fundamental patterns in tone and language don’t change over time. An attempt at solicitation is always an attempt at solicitation, and will always bear common characteristics.

For this reason, this approach only plays one small part of a very large engine. It gives an additional indication about the nature of the threat, but is not in itself used to determine anomalous emails.

Detecting the unknown unknowns

In addition to using the above approach to identify intention, Darktrace uses unsupervised machine learning, which starts with extracting and extrapolating thousands of data points from every email. Some of these are taken directly from the email itself, while others are only ascertainable by the above intention-type analysis. Additional insights are also gained from observing emails in the wider context of all available data across email, network and the cloud environment of the organization.

Only after having a now-significantly larger and more comprehensive set of indicators, with a more complete description of that email, can the data be fed into a topic-indifferent machine learning engine to start questioning the data in millions of ways in order to understand if it belongs, given the wider context of the typical ‘pattern of life’ for the organization. Monitoring all emails in conjunction allows the machine to establish things like:

  • Does this person usually receive ZIP files?
  • Does this supplier usually send links to Dropbox?
  • Has this sender ever logged in from China?
  • Do these recipients usually get the same emails together?

The technology identifies patterns across an entire organization and gains a continuously evolving sense of ‘self’ as the organization grows and changes. It is this innate understanding of what is and isn’t ‘normal’ that allows AI to spot the truly ‘unknown unknowns’ instead of just ‘new variations of known bads.’

This type of analysis brings an additional advantage in that it is language and topic agnostic: because it focusses on anomaly detection rather than finding specific patterns that indicate threat, it is effective regardless of whether an organization typically communicates in English, Spanish, Japanese, or any other language.

By layering both of these approaches, you can understand the intention behind an email and understand whether that email belongs given the context of normal communication. And all of this is done without ever making an assumption or having the expectation that you’ve seen this threat before.

Years in the making

It’s well established now that the legacy approach to email security has failed – and this makes it easy to see why existing recommendation engines are being applied to the cyber security space. On first glance, these solutions may be appealing to a security team, but highly targeted, truly unique spear phishing emails easily skirt these systems. They can’t be relied on to stop email threats on the first encounter, as they have a dependency on known attacks with previously seen topics, domains, and payloads.

An effective, layered AI approach takes years of research and development. There is no single mathematical model to solve the problem of determining malicious emails from benign communication. A layered approach accepts that competing mathematical models each have their own strengths and weaknesses. It autonomously determines the relative weight these models should have and weighs them against one another to produce an overall ‘anomaly score’ given as a percentage, indicating exactly how unusual a particular email is in comparison to the organization’s wider email traffic flow.

It is time for email security to well and truly drop the assumption that you can look at threats of the past to predict tomorrow’s attacks. An effective AI cyber security system can identify abnormalities with no reliance on historical attacks, enabling it to catch truly unique novel emails on the first encounter – before they land in the inbox.

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.
Author
Dan Fein
VP, Product

Based in New York, Dan joined Darktrace’s technical team in 2015, helping customers quickly achieve a complete and granular understanding of Darktrace’s product suite. Dan has a particular focus on Darktrace/Email, ensuring that it is effectively deployed in complex digital environments, and works closely with the development, marketing, sales, and technical teams. Dan holds a Bachelor’s degree in Computer Science from New York University.

Book a 1-1 meeting with one of our experts
Share this article

More in this series

No items found.

Blog

/

OT

/

April 4, 2025

Darktrace Named as Market Leader in the 2025 Omdia Market Radar for OT Cybersecurity Platforms

Default blog imageDefault blog image

We are pleased to announce that Darktrace / OT has been named a Market Leader in Omdia’s  2025 Market Radar for OT Cybersecurity Platforms. We believe this highlights our unique capabilities in the OT security market and follows similar recognition from Gartner who recently named Darktrace / OT as the sole Visionary in in the Magic Quadrant for Cyber Physical Systems (CPS) Protection Platforms market.

Historically, IT and OT systems have been managed separately, creating challenges due to the differences of priorities between the two domains. While both value availability, IT emphasizes confidentiality and integrity whereas OT focuses on safety and reliability. Organizations are increasingly converging these systems to reap the benefits of automation, efficiency, and productivity (1).

Omdia’s research highlights that decision makers are increasingly prioritizing comprehensive security coverage, centralized management, and advanced cybersecurity capabilities when selecting OT security solutions (1).

Rising productivity demands have driven the convergence of OT, IT, and cloud-connected systems, expanding attack surfaces and exposing vulnerabilities. Darktrace / OT provides a comprehensive OT security solution, purpose-built for critical infrastructure, offering visibility across OT, IoT, and IT assets, bespoke risk management, and industry-leading threat detection and response powered by Self-Learning AITM.

Figure 1: Omdia vendor overview for OT cybersecurity platforms
Figure 1: Omdia vendor overview for OT cybersecurity platforms

An AI-first approach to OT security  

Many OT security vendors have integrated AI into their offerings, often leveraging machine learning for anomaly detection and threat response. However, only a few have a deep-rooted history in AI, with longstanding expertise shaping their approach beyond surface-level adoption.

The Omdia Market Radar recognizes that Darktrace has extensive background in the AI space:

“Darktrace has invested extensively in AI research to fuel its capabilities since 2013 with 200-plus patent applications, providing anomaly detection with a significant level of customization, helping with SOC productivity and efficiency, streamlining to show what matters for OT.” (1)

Unlike other security approaches that rely on existing threat data, Darktrace / OT achieves this through Self-Learning AI that understands normal business operations, detecting and containing known and unknown threats autonomously, thereby reducing Sec Ops workload and ensuring minimal downtime

This approach extends to incident investigations where an industry-first Cyber AI AnalystTM automatically investigates all relevant threats across IT and OT, prioritizes critical incidents, and then summarizes findings in an easily understandable view—bringing production engineers and security analysts together to communicate and quickly take appropriate action.

Balancing autonomous response with human oversight

In OT environments where uptime is essential, autonomous response technology can be approached with apprehension. However, Darktrace offers customizable response actions that can be set to “human confirmation mode.”

Omdia recognizes that our approach provides customizable options for autonomous response:

“Darktrace’s autonomous response functionality enforces normal, expected behavior. This can be automated but does not need to be from the beginning, and it can be fine-tuned. Alternative step-by-step mitigations are clearly laid out step-by-step and updated based on organizational risk posture and current level of progress.” (1)

This approach allows security and production to keep humans-in-the-loop with pre-defined actions for potential attacks, enforcing normal to contain a threat, and allowing production to continue without disruption.  

Bespoke vulnerability and risk management

In the realm of OT security, asset management takes precedent as one of the key focus points for organizations. With a large quantity of assets to manage, practitioners are overwhelmed with information with no real way to prioritize or apply them to their unique environment.

Darktrace / OT is recognized by Omdia as having:

“Advanced risk management capabilities that showcase metrics on impact, exploit difficulty, and estimated cost of an attack […] Given the nascency of this capability (April 2024), it is remarkably granular in depth and insight.” (1)

Enabling this is Darktrace’s unique approach to AI extends to risk management capabilities for OT. Darktrace / OT understands customers’ unique risks by building a comprehensive and contextualized picture that goes beyond isolated CVE scoring. It combines attack path modeling with MITRE ATT&CK  techniques to provide hardening recommendations regardless of patching availability and gives you a clearer view of the potential impact of an attack from APT groups.

Modular, scalable security for industrial environments

Organizations need flexibility when it comes to OT security, some want a fully integrated IT-OT security stack, while others prefer a segregated approach due to compliance or operational concerns. The Darktrace ActiveAI Security Platform offers integrated security across multiple domains, allowing flexibility and unification across IT and OT security. The platform combines telemetry from all areas of your digital estate to detect and respond to threats, including OT, network, cloud, email, and user identities.

Omdia recognizes Darktrace’s expansive coverage across multiple domains as a key reason why organizations should consider Darktrace / OT:

“Darktrace’s modular and platform, approach offer’s integrated security across multiple domains. It offers the option of Darktrace / OT as a separate platform product for those that want to segregate IT and OT cybersecurity or are not yet in a position to secure both domains in tandem. The deployment of Darktrace’s platform is flexible—with nine different deployment options, including physical on-premises, virtual, cloud, and hybrid.” (1)

With flexible deployment options, Darktrace offers security teams the ability to choose a model that works best for their organization, ensuring that security doesn’t have to be a “one-size-fits-all” approach.

Conclusion: Why Darktrace / OT stands out in Omdia’s evaluation

Omdia’s 2025 Market Radar for OT Cybersecurity Platforms provides a technical-first, vendor-agnostic evaluation, offering critical insights for organizations looking to strengthen their OT security posture. Darktrace’s recognition as a Market Leader reinforces its unique AI-driven approach, flexible deployment options, and advanced risk management capabilities as key differentiators in an evolving threat landscape.

By leveraging Self-Learning AI, autonomous response, and real-world risk analysis, Darktrace / OT enables organizations to detect, investigate, and mitigate threats before they escalate, without compromising operational uptime.

Read the full report here!

References

  1. www.darktrace.com/resources/darktrace-named-a-market-leader-in-the-2025-omdia-market-radar-for-ot-cybersecurity-platforms
Continue reading
About the author
Pallavi Singh
Product Marketing Manager, OT Security & Compliance

Blog

/

Cloud

/

April 2, 2025

Fusing Vulnerability and Threat Data: Enhancing the Depth of Attack Analysis

Default blog imageDefault blog image

Cado Security, recently acquired by Darktrace, is excited to announce a significant enhancement to its data collection capabilities, with the addition of a vulnerability discovery feature for Linux-based cloud resources. According to Darktrace’s Annual Threat Report 2024, the most significant campaigns observed in 2024 involved the ongoing exploitation of significant vulnerabilities in internet-facing systems. Cado’s new vulnerability discovery capability further deepens its ability to provide extensive context to security teams, enabling them to make informed decisions about threats, faster than ever.

Deep context to accelerate understanding and remediation

Context is critical when understanding the circumstances surrounding a threat. It can also take many forms – alert data, telemetry, file content, business context (for example asset criticality, core function of the resource), and risk context, such as open vulnerabilities.

When performing an investigation, it is common practice to understand the risk profile of the resource impacted, specifically determining open vulnerabilities and how they may relate to the threat. For example, if an analyst is triaging an alert related to an internet-facing Webserver running Apache, it would greatly benefit the analyst to understand open vulnerabilities in the Apache version that is running, if any of them are exploitable, whether a fix is available, etc. This dataset also serves as an invaluable source when developing a remediation plan, identifying specific vulnerabilities to be prioritised for patching.

Data acquisition in Cado

Cado is the only platform with the ability to perform full forensic captures as well as utilize instant triage collection methods, which is why fusing host-based artifact data with vulnerability data is such an exciting and compelling development.

The vulnerability discovery feature can be run as part of an acquisition – full or triage – as well as independently using a fast ‘Scan only’ mode.

Figure 1: A fast vulnerability scan being performed on the acquired evidence

Once the acquisition has completed, the user will have access to a ‘Vulnerabilities’ table within their investigation, where they are able to view and filter open vulnerabilities (by Severity, CVE ID, Resource, and other properties), as well as pivot to the full Event Timeline. In the Event Timeline, the user will be able to identify whether there is any malicious, suspicious or other interesting activity surrounding the vulnerable package, given the unified timeline presents a complete chronological dataset of all evidence and context collected.

Figure 2: Vulnerabilities discovered on the acquired evidence
Figure 3: Pivot from the Vulnerabilities table to the Event Timeline provides an in-depth view of file and process data associated with the vulnerable package selected. In this example, Apache2.

Future work

In the coming months, we’ll be releasing initial versions of highly anticipated integrations between Cado and Darktrace, including the ability to ingest Darktrace / CLOUD alerts which will automatically trigger a forensic capture (as well as a vulnerability discovery) of the impacted assets.

To learn more about how Cado and Darktrace will combine forces, request a demo today.

Continue reading
About the author
Paul Bottomley
Director of Product Management, Cado
Your data. Our AI.
Elevate your network security with Darktrace AI