A Budding Criminologist’s Perspectives On OSINT
Author’s Introduction: Olivia Elliott is a second-year student at Virginia Tech, pursuing a bachelor’s degree with a major in criminology and double minoring in leadership studies and sociology. She is also pursuing an Honors Laureate Diploma through the Virginia Tech Honors College and is a proud member of the Virginia Tech Corps of Cadets. Additionally, Olivia is currently an intern at Plessas Experts Network, Inc., hoping to gain experience that will prove valuable to a future career in the federal government. In her spare time, Olivia enjoys hanging out with her friends and family, cuddling with her dogs, and crocheting while listening to audiobooks and podcasts.
A Budding Criminologist’s Perspectives On OSINT
TW: Readers are advised that the following content may be disturbing. Please read with caution.
My natural desire to discover the truth led me to apply as a criminology major at Virginia Tech and start learning more about open-source intelligence (OSINT). I recognize that only some in the OSINT community have a criminological background, so OSINT may relate to different ideas or concepts to them than it does to me. Due to my lack of experience in other fields, I cannot describe their perspectives, but I can offer a look at OSINT from a criminological lens.
Important Terminology
OSINT
First, I’ll need to define a couple of terms so that everyone understands what I’m talking about here. For those new to OSINT, we’ll first need to explain the concept of open-source intelligence. The book Deep Dive by Rae Baker contains a wonderful definition, which is as follows: “Open-source intelligence (OSINT) is the production of intelligence through the collection and enrichment of publicly available information.” Baker adds that "[...] OSINT is a purely passive method of intelligence collection, meaning that we view information such as a person’s credentials in a database, but we do not use those credentials to access anything or to log in.” In plain terms, OSINT means collecting, analyzing, and applying data obtained without hacking, using passwords, etc., to solve a problem or answer a question.
However, just because OSINT is publicly available, it doesn’t mean there aren’t restrictions. Baker notes that OSINT investigations have rules just the same as any regular type of investigation, so if you are just starting to get into OSINT, it’s critical to review the legislation in your area (Baker cautions that both national and state guidelines are necessary to look into) so that you know that whatever evidence you pull will be useful and so that you don’t end up in legal trouble or lose your job. If, for some reason, one type of OSINT isn’t available to you, never fear! Deep Dive notes that there are many different types of OSINT, each corresponding to various kinds of data, ranging from numerical data to EXIF data to geolocation data.
Criminology and Sociology
It’s also essential to define criminology, but before that, it is crucial to understand the meaning of sociology. The University of North Carolina at Chapel Hill (UNC) says that “Sociology is the study of human social relationships and institutions,” but also that it “[...] is an exciting and illuminating field of study that analyzes and explains important matters in our personal lives, our communities, and the world." To summarize, it is the study of humans interacting with each other, our surroundings, and the ideas present in the world.
Criminology is actually an offshoot of sociology, which is both why I explained sociology first and why so many criminology and sociology majors share classes, at least in their first years of college. It’s also common to double major or minor in the one you aren’t majoring in. But, you might ask, where is the difference? North Central College defines criminology as follows: “Criminology refers to the study of crime and criminals within a societal sphere.” Essentially, this means that instead of focusing on human behavior in general, we are specifically looking at criminal behavior and the factors that contribute to that behavior. In this article, I will use the experiences gained from utilizing this lens to describe my perception of OSINT.
Further terms will need to be defined to better understand this perception, but these are predominantly in the criminology/sociology (CRIM/SOC) field and may be less common to readers, so their explanation is even more critical.
Deviance
The first is pretty simple: deviance, which, funnily enough, I first learned about in a course called Deviant Behavior. Deviance is going against (deviating, if you will) from societal norms. Deviant behavior doesn’t always come in the form of committing a crime. Say you and your family attend a wedding in the United States (I clarify the setting because traditions vary in different locales). The bride wears a gorgeous white gown as is the traditional custom. But what’s this?! Your Aunt Sheila has also decided to wear white. Everyone starts muttering when she walks in, and throughout the reception, nobody talks to her. Or if they do, it’s in curt or passive-aggressive tones. Why? Because while Sheila hasn’t committed a crime written in the legal code, she has, in a way, committed a crime of etiquette.
The same can be said of talking loudly in a movie theater, picking your nose, or wearing an ‘eccentric’ outfit. As you can probably tell, there are varying degrees of deviance. The difference between each is that the more outside the norm the behavior is, the more people will take offense to it, laugh at it, or otherwise punish the person doing it. Regardless of whether the behavior does any harm or not, if it is not considered socially acceptable, the person is being deviant.
Strain
Another thing we need to cover is the concept of strain, which I have studied in many classes. Multiple theories by varying scientists talk about the phenomenon in which different societal factors come together and put pressure, or strain, on a person, which may ultimately lead to them committing a crime, sometimes because they can’t imagine another possible option. This is referenced on Britannica’s website and expounds on the different theories associated with the topic. Specifically, there is a theory mentioned called Merton’s Strain theory. In that theory, it is suggested that all people are trying to reach goals that are commonly agreed upon in their shared locale (known as the common goals). For example, many people in the United States are trying to be well-off financially and attain things that give them opportunities, such as a college education.
However, most people also try to accomplish these goals through the accepted means. This means that people generally attempt to achieve their goals in ways considered normative to others and aren’t against the law. But, if certain societal factors or prejudices present an obstacle, people may abandon the common goals, accepted means, or both altogether. Below is a chart showing the different demographics of people who undergo strain. You’ll note that the conformists accept both the common goals and accepted means. They actually don’t undergo strain. This is what everyone, at least initially, strives for. Strain theory can also apply to non-criminal actions or behaviors that showcase minor deviance. That’s my belief, anyhow, and this idea of mine will heavily come into play later on in this article.
Sourced from: Twitter/X (@allsociology)
Stigma
Lastly, it’s vital to understand the concept of stigma (another term I learned in Deviant Behavior. The National Library of Medicine offers a wonderful definition, saying that “[...] stigma [is] a social attribute that is discrediting for an individual or group." If you want more information on stigma, another great resource is Erving Goffman’s book, Stigma, in which he first introduced the concept. To summarize, stigma is a trait of one or many people that is deviant or that most of society finds unlikeable.
Now that you know the relevant terms, we can connect them to OSINT. I learned most of my OSINT skills from an OSINT 101 course taught by Plessas Experts Network, so most concepts and ideas relating directly to OSINT come from there. It’s a great course, and I highly recommend it for anyone wanting to learn more about any of the following topics and many more that I regrettably do not have time to list here. But, I will say that one of the topics we covered was social media use in OSINT investigations, as well as its general history and the demographics associated with different platforms, and that is a prime topic to highlight the connection between OSINT and criminology.
Social Media
Connection to Strain
Firstly, social media is connected to my criminological experience through the prevalence of strain on each platform. I would go so far as to say that strain is the foundation of social media. Everyone looking to be a part of the conformist group that follows the common goals and uses the accepted means knows that Americans (I can’t speak for other countries, but I imagine it’s the same in many countries that have high numbers of social media users) want to know and emulate what is trendy at all times, which will allow them to attain the common goal of being famous and well-liked. Because social media is becoming increasingly image-heavy, it makes it much easier to take in larger amounts of information on what’s in than there was initially with text-heavy social media. So people are frantically posting anything and everything about their lives, trying to please the rest of the world (or at least pander to a specific group) by copying whatever new style or dance is popular. This has made social media an effective, accepted means of achieving the common goal of popularity, which we can see with the sheer number of influencers and people making money solely from their social media content. There could be no possible downsides to posting so much, right?
OPSEC Concerns
However, as many people in the OSINT community will tell you, the more you post, the more potential there is for danger. I’m not saying you should be afraid to use social media; it is an incredible invention that has a lot of positive impacts, both in and out of the investigative community. But, it’s also important to be cautious about what you post. You might get so caught up in copying the latest trend that you accidentally include information online predators can exploit and use to take advantage of you. If you take a picture of yourself in front of your house and post it online, it’s incredibly easy for people to take that image and put it into Google Image Search, and, boom, now the entire Internet has the potential to know where you live. Due to the incredibly easy nature of information sharing on social media, people (especially those unfamiliar with social media or operational security (OPSEC)) are more likely to post information that gives away sensitive information accidentally. However, this isn’t the only potential concern associated with excessive social media use.
Parasocial Relationships
One should also be cautious of parasocial relationships, a concept I didn’t learn about until very recently and which I found a fascinating sociological phenomenon. Verywell Mind defines parasocial relationships as “[...] a one-sided relationship that a media user engages in with a media persona” of any type, real or fictional. Particularly important is a nested concept known as parasocial interactions, which, as VeryWell Mind also states, “[...] take place exclusively while interacting with a persona via media and psychologically resemble real-life face-to-face interactions."
The article goes on to state that we’ve become so hardwired as a species to register face-to-face contact as social bonding, and that video technology is so new a technology that we haven’t evolved subconsciously enough to think of seeing people on social media (or any other media, for that matter) as anything other than being around them in real life. This is great because it can help us feel less alone and more connected, but just as connection and camaraderie manifest themselves in media-based and real-world interactions, so do heartbreak and animosity. This leads to parasocial attachments and parasocial breakups, respectively. And while ‘attachment’ may sound like a positive term (and, in the case of instances like celebrity crushes, it can), it can approach extremes.
For example, there was a documentary I watched once for Deviant Behavior that was called Beware the Slenderman. It discusses an attempted murder that took place when two girls became obsessed with the Slenderman, a horror figure described “[...] as an abnormally tall, thin man, with a featureless face” in an article by CBS. Between this fanaticism that stemmed from horror content on the Internet and some mental health problems, the girls became convinced that the Slenderman needed them to kill another one of their friends, and so they took her into the woods one day and stabbed her multiple times, which, thankfully, she survived. The presence of social media makes this phenomenon concerning because platforms like TikTok and Instagram allow users to consume a lot of content in a short span of time, potentially allowing them to forge parasocial relationships more readily, and so more incidents like what happened in Beware the Slenderman may occur with others who want, as the synopsis for the documentary on IMDb says, “[...] to appease [...]” the characters or people they are in parasocial relationships with.
The Allure of Quantitative Data
Speaking of appeasement, we discuss, in many of my data analytics and statistics classes, that numbers make humans happy. They make things simple, neat, and measurable. Social media provides this need for information, which is particularly important because people want to ensure they meet the goal of being cool and popular. Users can satisfy their curiosity on this front by simply checking how many people have liked their posts, how many followers they have, and how many people have commented. There is constant competition in today’s world for users (particularly in the young adult range) to have the highest (or, at the very least, respectable) amount of likes, followers, snaps, etc. This can present a problem, especially if someone has fewer followers or likes or has comments on their posts that are negative or attacking. Consequently, a person will develop low self-esteem and may even go so far as becoming depressed or suicidal, which has led to cyberbullying being highlighted more and more as social media makes more and more of an impact on today’s youth.
Connection to Deviance
The concept of deviance can also play a role in social media, at least in the etiquette sense, because, as we all know, different social media platforms have specific ways that you’re ‘meant’ to use them. For instance, Facebook is used to connect with others, particularly family and friends, and to share pictures or videos with stories or memories posted with them, which are meant to be treasured forever. Snapchat, on the other hand, is a more casual platform that is mainly used for sharing quick videos or pictures that people don’t attach much emotional meaning to or want to be discreet and temporary since the posts only stay up for a limited amount of time (which, by the way, is why drug dealers commonly use Snapchat to communicate. You wouldn’t want to post it on Facebook and have a higher risk of people becoming aware of your activities; that would be considered deviant within the criminal community).
There’s also a generational aspect of deviance within social media. You’ll notice that most people who post on Instagram, Snapchat, or TikTok are relatively young. In contrast, people really only start to seriously use Facebook or Twitter (now called X) when they are in their adult or elder years. If you were to see a video on Instagram of your friend’s grandma doing a dance trend to a popular rap song and posting it on a personal account, that would be considered deviant, both because elderly people don’t usually dance to rap music (or do dances in the contemporary style, for that matter) or post on Instagram. Knowing these demographics can be helpful in OSINT investigations, which, coincidentally, is the next topic.
OSINT Investigations
Are Investigators Deviant?
Deviance can, in addition to relating to social media, be tied to OSINT investigation. Since the latest technology (social media, for example) is new and still somewhat out of the norm due to its exponentially evolving nature, it is deviant, making the data stored within inherently deviant. In addition, due to the developments in new technology, there have also been updates in the tools that investigators use to solve crimes. Thus, investigations have also become deviant because people still perceive the investigative norm as mainly focusing on tangible evidence. While physical investigation techniques will never be fully redundant, digital investigation, including the majority of commonly used OSINT, is becoming increasingly common. However, the public perception of investigation has yet to catch up to this reality, so OSINT is deviant, at least for now, in more ways than one.
Deviant Amateurs: Asset or Liability?
There’s another reason OSINT investigations are deviant: amateurs and people not in official investigative positions can join in on investigations. A couple of noteworthy examples come to mind. The first I learned of when researching this article was the story of the group Bellingcat.
An article from France24 gives a pretty good overview of what the group does. Essentially, they act as investigators into the doings of various governments because nobody within international agreements, or individual countries, for that matter, is, in specific issues, willing to do it themselves, especially when it will create a scandal. Granted, investigations aren’t entirely outside the sphere of their everyday jobs (some of them are journalists), but I feel like most journalists don’t go to the lengths that Belligcat has. France24 notes that among their accomplishments has been the uncovering of various recent misdeeds and sabotages in the Russian government, although the Russian government is by no means Bellingcat’s particular target. The article says that one of the group members, when interviewing him, “[...] cite[d] investigations into the Syrian war, EU police agency Europol and others focused on Greece, Turkey, Hungary and the far right in the United States and Europe."
I suppose many journalists are used to this sort of danger (although these ones are treading a very fine line), but the average Joe certainly isn’t, and that’s where the second example comes into play. In my Deviant Behavior class, we had to watch a three-episode documentary called Don’t F**k With Cats. It is about a group of people on the internet who became enraged at a video that was posted on Facebook that, I am sorry to say, contained the brutal killing of a kitten in a context that indicated that it was for the personal enjoyment of the killer. Some people, horrified by what they saw, banded together and tried to figure out who the person who had done such a terrible deed was. As they searched, finding more and more information as they went along, the perpetrator posted more videos of kitten murder, escalating in their brutality until, finally, he went a step further and committed homicide. Eventually (spoiler alert), they caught the killer, with critical evidence from the group of amateurs used to ultimately confirm his guilt and find his whereabouts.
However, in doing so, the group (as it appears in the docuseries; I know the series doesn’t tell the whole story, and likely, I may be biased) seemed to dive right in without taking appropriate precautions. The way they describe it, they got so enraged that they decided to investigate immediately. While one of the group members had a fake account they were investigating Facebook under, the perpetrator still discovered her identity and found out where she worked. OPSEC is critical for investigators, and while amateurs can be incredibly useful, they don’t know everything, especially not how to keep themselves safe while investigating (I include myself in this assessment, recognizing that I am an amateur myself).
Additionally, in the process of their investigations, the group trying to identify the kitten killer (prior to the homicide) at first thought they had found the culprit due to a self-declared boast of guilt, but it turned out to be an internet troll. However, before anyone found this out, they had already threatened the person trolling the group, calling him names and bullying him. Eventually, he committed suicide. None of the group meant for it to happen, but investigations sometimes go very, very wrong. Since none of the amateurs (at least, none mentioned outright in the docuseries) were official investigators, they didn’t have the experience or training to know that one always needs to fully confirm everything and not jump to conclusions.
Again, amateurs do still have great things to offer to OSINT investigations. They think of ways to glean information in plain sight, while official investigators have kind of been trained to rely on specific databases and exclusive tools. However, when doing OSINT, or any investigation, really, you need to have a firm understanding of the responsibility you hold to find the absolute truth and to do so without endangering yourself or anyone else in the process, and this is, unfortunately, where many amateurs struggle. Another area of struggle can be distinguishing truth from fact and avoiding bias in investigations, which I will discuss next.
Informational Echo Chambers, Source Reliability, and Deepfaking
In at least one of my CRIM/SOC courses, we talked about informational echo chambers. This common phenomenon consists of a person only seeking out and trusting information sources that say things that they agree with. If you’ve ever wondered why some people prefer one news channel over another, this idea explains why.
Besides the risk of biases affecting the judgment of media users, there’s also the risk that the content may be entirely fake. I’m not just talking about lies in the media, although that is, unfortunately, an ever-growing concern in today’s world. I’m talking about the development of technology that allows for the presentation of false information. During the Plessas Experts Network (PEN) OSINT 101 course, I was shocked to discover that anyone can open the inspect panel on their web browser to change pictures and words in a website to produce a fake display that can be screenshotted and shared to others. In OSINT 101, Kirby Plessas used the example of editing a news headline, demonstrating how easily someone can produce a very realistic screenshot despite the news media content being changed to something very different from the original article. If the user were to screenshot their work and send it to someone else, it would appear to be an official article and might be taken as believable.
And this isn’t even the technology with the most potential for sowing disinformation. The development of AI and deepfaking technology has become problematic for elections worldwide. One particularly troubling example described by multiple news articles (the best summary of the event can be found on NPR, but other good articles can be found through The Washington Post and The Hill) was the recent series of deepfake phone calls that, by all accounts, sounded like President Biden was delivering them. The calls consisted of faux-Biden attempting to persuade left-wing New Hampshire residents not to vote in the primary elections, saying, "‘Your vote makes a difference in November, not this Tuesday’ [...]." While this isn’t the most convincing example of a deepfake, people may still fall for it, and that is why the incident is being treated as a voter suppression case, notes NPR. OSINT knowledge can be an incredible superpower in cases like these. In PEN’s OSINT 101 class, we learned how to distinguish AI-generated images from real ones, a skill that will become useful when assessing source reliability during the election season.
Flawed Data
Data Biases
As scientists of all types, not just criminologists and sociologists, know, data and data collection are fickle things. When a concept is being studied or a hypothesis tested, no matter how objective the scientist attempts to be, there will always be some fraction of bias or personal opinion inserted into the study. Let’s say you’re studying poverty and its relationship to other variables. Poverty is brutal to measure because its definition is subjective. As I learned in my Community Analytics course, people who study poverty have many ways of quantifying it, so findings will vary from study to study. This limits OSINT investigations because investigators rely on public data, meaning they may pick up unreliable information.
Flaws in AI and Geolocation
This is particularly problematic with AI because the sources that it is primed to pull information from are not always accurate or may be biased. While it sometimes produces correct output, we cannot check its work because chatbots do not list its sources. AI can read all the biases and false information on the Internet, so OSINT investigators must be careful when using it as a research tool.
Geolocation is also somewhat problematic. In high school, I took a couple of Geospatial Information Sciences (GIS) courses and learned how to do a lot of research that way. The tool we used most often was called ArcGIS, although we also used ArcGISPro a lot as well. Both resources had vast databases of data collected by many organizations, which we would use to investigate relationships. One project we did was analyzing Chicago’s crime data for a set time. However, I later learned that crime data has a huge problem: it’s wholly inaccurate. Don’t get me wrong, crime data is essential, but the problem is that it depends on equitable policing. As I’ve learned in many criminology courses, there are inherent biases that people have against other groups of people, making them racist, sexist, homophobic, classist, etc., without them being consciously aware of it. I’m not excusing this behavior by saying everyone is affected by it; quite the opposite. I’m saying that because we have these biases, we need to be aware of them when collecting data. Unfortunately, because people who collect crime data (law enforcement, investigators, scientists, etc.) are human, they are biased on where they look. That’s why there is a disproportionate number of arrests in poor neighborhoods, especially in those mostly populated with people of color. But what does this have to do with GIS? One incredibly cool feature of ArcGIS is that it can create cluster or hot spot maps, showing high concentrations and low concentrations of data points, making them a go-to tool for people who want to predict future data.
However, because crime data is inherently biased, we can already predict, based on our stereotypes of crime, that there will be a higher concentration in poor, urban neighborhoods. The reality is that rich people, poor people, races of all types, and people of all genders commit crimes roughly about the same amount. But, because there is simply not enough data being collected in rural or suburban White neighborhoods, we get a disproportionate, misrepresentative view of the data, which is detrimental to other researchers and investigators. And, because people will mistake the output for the truth, policing will continue to revolve around poor, urban neighborhoods where there will be even more data collection, ‘confirming’ the findings of police and investigators.
Conclusion
I am grateful to have been given the idea to write this article, as I hope it will positively impact the OSINT community. This network of truth-seekers will only thrive if we are willing to share our unique stories and perspectives based on our diverse backgrounds. Our differences are assets, allowing us to see the world differently. When we combine these ways of seeing OSINT, we can better understand the intricacies of social media culture and its differences from other technology. We can more readily accept different viewpoints highlighting the implications of our work if we discuss them. Lastly, when we converse about OSINT tools and data, we will be privy to various opinions on their merits and dangers, particularly as they pertain to using tools in the media and how data may or may not be reliable. I hope sharing my perspectives will foster connections between OSINT community members from diverse backgrounds. Everyone has a particular set of unique skills, and I hope this article will spark discussions that will bring those to light. To help start the ball rolling, feel free to reach out to me on LinkedIn and share your insights!
Resource List
Editing Assisted by
ChatGPT
Grammarly
Kyle Elliott, COO, Plessas Experts Network, Inc.
Kirby Plessas, CEO, Plessas Experts Network, Inc.