How writing patterns, online activities, and other unintentional identifiers can be used in cyber offense and defense.

Kelly Sheridan, Former Senior Editor, Dark Reading

July 2, 2019

4 Min Read

As we move throughout our digital lives, we unknowingly leave traces — writing styles, cultural references, behavioral signatures — that can be compiled to form a profile of our online personas.

These identifiers are different from physical identifiers such as fingerprints, faces, handwriting, DNA, and voice, all of which allow law enforcement to trace crimes back to offenders and enable biometric authentication tools. But physical identifiers are often irrelevant when it comes to tracking criminals in the digital realm, where non-physical traits can prove useful.

Matt Wixey, head of technical research for PwC's cybersecurity practice in the UK, calls these behavioral identifiers "human side-channels" and says they often result from human actions. Human side-channels, he explains, are rooted in personality psychology and result from each individual's unique experiences, training, and feedback. We often don't know we exhibit them.

"These are ways you can be identified or tracked via unintentional or inadvertent leakage of behaviors," he says, adding that these traits are "predominantly unintentional."

While there are many behavioral traces to explore, Wixey chose forensic linguistics to explain how human side-channels can be used in offensive and defensive security. Linguistics is by no means a new discipline, but as a former law enforcement official-turned-cybersecurity researcher, he has found the study practical for investigating both physical and virtual crimes.

"It's kind of a spin-off of applied linguistics," he explains. "The principle is that everyone has a unique style of writing," not necessarily in terms of their handwriting but in how people construct sentences and paragraphs, as well as how they use punctuation and grammar. From a defensive perspective, this could be handy if a cybercriminal wrote any text in conjunction with the attack: a spearphishing email, for example, or a ransom note or text message.

Of course, the usefulness of a piece of text depends on how much is available to the investigator. "There are some things you can do just with the text itself," says Wixey, and when you have a piece of text, it's easier and more cost-effective to avoid a full forensic analysis. There are several ways security pros can leverage forensic linguistics, all of which vary depending on the resources they have and how prepared they are to invest in this technique.

As an example, let's say an organization was hit with a spearphishing attack. In the malicious email, analysts can seek unusual construction of sentences or stand-out phrases, and paste those into a search engine to see if they appear anywhere else on the Internet. This tactic has been used in real-world offenses, Wixey says, and it's a jumping-off point for further investigation.

If an unusual phrase appears in a recent forum post, it could prove useful to read through the forum for messages talking about the attack, or other clues that could provide more insight on what happened. The forum could also be passed to law enforcement as a possible lead.

Forensic linguistics can also be helpful in comparing social media accounts. If the same person operates multiple Twitter accounts, he adds, you may be able to tie both to one operator. This could prove useful in investigating disinformation campaigns or identifying extortion, fraud, or another psychological agenda.

With more time and resources, Wixey continues, a full-time attack investigator or threat intelligence analyst could compile a corpus, or collection, of text from different actors and sources. As they build a collection of ransom messages, tweets, and forum posts, they can compare future attacker texts to those in their repository and see if any matches exist.

"It's still below the radar in terms of most security practitioners' awareness," he says, adding that "it's just not most people's standard investigative protocols."

Writing in Disguise
Although these human side-channels are deeply ingrained into an individual's personal writing style, Wixey says there are ways people attempt to disguise them. With respect to forensic linguistics, they may run a text through Google Translate a dozen times and continuously tweak the text so its meaning is consistent, but the voice and structure are concealed. It's a "pretty primitive" strategy, he says, but it's also easy to automate. Another tactic is to collaborate with someone else on writing a piece of text so the two styles are scrambled.

At Black Hat USA, Wixey will examine multiple human side-channels, how they can be used in attacks and defense, privacy implications, and how they can be countered in his briefing, "I'm Unique, Just Like You: Human Side-Channels and Their Implications for Security and Privacy."

Related Content:

 

Black Hat USA returns to Las Vegas with hands-on technical Trainings, cutting-edge Briefings, Arsenal open-source tool demonstrations, top-tier security solutions, and service providers in the Business Hall. Click for information on the conference and to register.

Read more about:

Black Hat News

About the Author(s)

Kelly Sheridan

Former Senior Editor, Dark Reading

Kelly Sheridan was formerly a Staff Editor at Dark Reading, where she focused on cybersecurity news and analysis. She is a business technology journalist who previously reported for InformationWeek, where she covered Microsoft, and Insurance & Technology, where she covered financial services. Sheridan earned her BA in English at Villanova University. You can follow her on Twitter @kellymsheridan.

Keep up with the latest cybersecurity threats, newly discovered vulnerabilities, data breach information, and emerging trends. Delivered daily or weekly right to your email inbox.

You May Also Like


More Insights