Darktrace - Generative AI Changes Everything You Know About Email Cyber Attacks

Darktrace - Generative AI Changes Everything You Know About Email Cyber Attacks

In new data published today, Darktrace reveals that email security solutions, including native, cloud, and ‘static AI’ tools, take an average of thirteen days from an attack being launched on a victim to that attack being detected, leaving defenders vulnerable for almost two weeks if they rely solely on these tools.

In March 2023, Darktrace commissioned a global survey with Census-wide to 6,711 employees across the UK, United States, France, Germany, Australia and The Netherlands to gather third-party insights into human behavior around email, to better understand how employees globally react to potential security threats, their understanding of email security and the modern technologies that are being used as a tool to transform the threats against them.

Key findings (globally and U.S.) indicate:

  1. Eighty-two percent of global employees are concerned that hackers can use generative AI to create scam emails that are indistinguishable from genuine communication.
  2. The top three characteristics of communication that make employees think an email is a phishing attack are: being invited to click a link or open an attachment (68%), unknown sender or unexpected content (61%), and poor use of spelling and grammar (61%)
  3. Nearly 1 in 3 (30%) of global employees have fallen for a fraudulent email or text in the past
  4. Seventy percent of global employees have noticed an increase in the frequency of scam emails and texts in the last 6 months
  5. Eighty-seven percent of global employees are concerned about the amount of personal information available about them online that could be used in phishing and other email frauds
  6. Over a third of people have tried ChatGPT or other Gen AI chatbots (35%)

The Email Threat Landscape Today
Darktrace researchers observed a 135% increase in ‘novel social engineering attacks’ across thousands of active Darktrace/Email customers from January to February 2023, corresponding with the widespread adoption of ChatGPT. These novel social engineering attacks use sophisticated linguistic techniques, including increased text volume, punctuation, and sentence length with no links or attachments. The trend suggests that generative AI, such as ChatGPT, is providing an avenue for threat actors to craft sophisticated and targeted attacks at speed and scale.

In addition, threat actors are rapidly exploiting the news cycle to profit from employee fear, urgency, or excitement. The latest iteration of this is the collapse of Silicon Valley Bank (SVB) and the resulting banking crisis, which has presented an opportunity for attackers to spoof overly sensitive communication, for example seeking to intercept legitimate communication instructing recipients to update bank details for payroll. Seventy-three percent of employees working in financial services organizations have noticed an increase in the frequency of scam emails and texts in the last six months.

Innocent human error and insider threats remain an issue. Many of us (nearly 2 in 5) have sent an important email to the wrong recipient with a similar looking alias by mistake or due to autocomplete. This rises to over half (51%) in the financial services industry and 41% in the legal industry, adding another layer of security risk that is not malicious. A self-learning system can spot this error before the sensitive information is incorrectly shared

What Does the Arms Race for Generative AI Mean for Email Security?

Your CEO emails you to ask for information. It is written in the exact language and tone of voice that they typically use. They even reference a personal anecdote or joke. Darktrace’s research shows that 61% of people look out for poor use of spelling and/or grammar as a sign that an email is fraudulent, but this email contains no mistakes. The spelling and grammar are perfect, it has personal information, and it is utterly convincing. But your CEO did not write it. It was crafted by generative AI, using basic information that a cyber-criminal pulled from social media profiles.

The emergence of ChatGPT has catapulted AI into the mainstream consciousness - 35% of people have already tried ChatGPT or other Gen AI chatbots for themselves – and with it, real concerns have emerged about its implications for cyber defense. Eighty-two percent of global employees are concerned that hackers can use generative AI to create scam emails indistinguishable from genuine communications.

Emails from CEOs or other senior business leaders are the third highest type of email that employees are most likely to engage with, with over a quarter of respondents (26%) agreeing. Defenders are up against Generative AI attacks that are linguistically complex and entirely novel scams that use techniques and reference topics that we have never seen before. In a world of increasing AI-powered attacks, we can no longer put the onus on humans to determine the veracity of communications. This is now a job for artificial intelligence.

By understanding what is normal, it can determine what does not belong in a particular individual’s inbox. Email security systems get this wrong too often, with 79% of respondents saying that their company’s spam/security filters incorrectly stop important legitimate emails from getting to their inbox.

With a deep understanding of the organization, and how the individuals within it interact with their inbox, the AI can determine for every email whether it is suspicious and should be actioned or if it is legitimate and should remain untouched.

This approach can stop threats like:

  1. Phishing
  2. CEO Fraud
  3. Business Email Compromise (BEC)
  4. Invoice Fraud
  5. Phishing scams
  6. Data Theft
  7. Social Engineering
  8. Ransomware & Malware
  9. Supply Chain Attack:
  10. URL-based spear-phishing
  11. Account Takeover
  12. Human Error
  13. Ransomware & Malware
  14. Microsoft
  15. Insider Threat

Self-learning AI in email, unlike all other email security tools, is not trained on what ‘bad’ looks like but instead learns you and the normal patterns of life for each unique organization.

Social engineering – specifically malicious cyber campaigns delivered via email - remain the primary source of an organization’s vulnerability to attack. Popularized in the 1990s, email security has challenged cyber defenders for almost three decades. The aim is to lure victims into divulging confidential information through communication that exploits trust, blackmails or promises reward so that threat actors can get to the heart of critical systems.

Social engineering is a profitable business for hackers – according to estimates, around 3.4 billion phishing e-mails get delivered every day.

As organizations continue to rely on email as their primary collaboration and communication tool, email security tools that rely on knowledge of past threats are failing to future-proof organizations and their people against evolving email threats.

Widespread accessibility to generative AI tools, like ChatGPT, as well as the increasing sophistication of nation-state actors, means that email scams are more convincing than ever.

Humans can no longer rely on their intuition to stop hackers in their tracks; it is time to arm organizations with an AI that knows them better than attackers do.


Featured Cybersecurity


New Products

  • LiftMaster Garage Door Opener

    LiftMaster Garage Door Opener

    LiftMaster Transforms the Garage Door Opener Into a Sleek Smart Home Device That Does More Than Open and Close the Garage Door 3

  • QCS7230 System-on-Chip (SoC)

    QCS7230 System-on-Chip (SoC)

    The latest Qualcomm® Vision Intelligence Platform offers next-generation smart camera IoT solutions to improve safety and security across enterprises, cities and spaces. The Vision Intelligence Platform was expanded in March 2022 with the introduction of the QCS7230 System-on-Chip (SoC), which delivers superior artificial intelligence (AI) inferencing at the edge. 3

  • Unique Oversized ID Card Printer

    Unique Oversized ID Card Printer

    Idesco Corp. is announcing its card printer – the XCR100 2.0 printer- that allows customers to personalize oversized ID cards on demand. The printer is ideal for assisting healthcare organizations find the right badging solution. As healthcare facilities continue to combat the spread of COVID-19, issuing oversized ID cards has helped identify staff clearly while adding an extra layer of security. The XCR100 2.0 printer is the only dye-sublimation printer on the market that can personalize CR100 cards (3.88" x 2.63"). The cards that are 42% larger than the standard credit card size. The printer can produce up to 180 full cards per hour in color, and up to 1,400 cards per hour in monochrome. An optional flipper is available to print dual-sided badges in one pass. Contactless encoding comes as an option to help healthcare facilities produce secure access badges on demand and the card printer features a 2-year warranty. 3