Colin Gray on How ight we Create a Common Language on Dark Patterns.

Dark patterns go against most UX design principles, yet they are pretty much everywhere. So how is this huge prevalence explained? What is missing in UX designer training to avoid deceptive designs? Is the regulation sufficient to deter companies from using dark patterns? How might we create a common language and hierarchy of dark patterns that aligns differing terminology from scholars and regulators?

To discuss the topic, Marie Potel speaks with Colin Gray, Associate Professor at the Luddy School of Informatics, Computing and Engineering at Indiana University Bloomington and Director of the human computer interaction design program.

To go further:

• "Towards a preliminary ontology of dark patterns knowledge", by Colin Gray, Cristina Santos, Nataliia Bielova
  

Have a question or need some support? Visit us at fairpatterns.com and follow us on LinkedIn: FairPatterns!
‍

‍
Transcript

Marie Potel-Saville 

Hi everyone, and welcome to this new episode of fighting dark patterns, regain your freedom online. My name is Marie Potel-Saville and I'm the founder of fair patterns, the first holistic solution to detect dark patterns, remedy them with interfaces that empower users to make enlightened and free choices, and train all stakeholders to avoid creating new Dark patterns or any deceptive design. As you probably know by now, listening to this podcast, the so called Dark patterns are interfaces that deceive or manipulate users to make you behave either without realizing or even against your interests. They're also called deceptive design or deceptive design patterns and unfortunately, they’re pretty much everywhere. It's like a sort of digital plague. For example, the European Commission identified last year in 2022, that there is at least one dark pattern in 97% of European favorite e-commerce websites. Now, one of the interesting things about dark patterns is the amazing volume of research since the term was coined in 2012 by Harry Brignull. Literally hundreds of scientific articles have been published over the past decade all over the world and we had the pleasure to analyze all of these articles in our r&d lab. So there are hundreds of different types of dark patterns which have been identified in scientific literature with interesting names such as the famous Roach motel : easy to subscribe very difficult or even impossible to unsubscribe, or Confirm shaming : making users feel ashamed of clicking on a different button. To date there are no less than 16 Different taxonomies created by scholars or regulators such as the Federal Trade Commission, the competition and market authority, or the French Data Protection Authority. So basically, for anyone trying to address the topic, whether you're a designer, or developer, a marketer, a business owner, a regulator, the first question is probably, what exactly are we talking about? And that's precisely why I'm so happy to talk to Colin Gray today as a co-author of the first ontology on dark patterns, to talk with him about how might we create a common language on dark patterns. Hi, Colin !

‍

Colin M. Gray 

Very nice to join you all. Thank you for the invitation.

‍

Marie Potel-Saville

It's really my pleasure Colin, thanks a lot. So Colin, you are an associate professor in the School of Informatics Computing and Engineering at Indiana University Bloomington, where you are director of the Human Computer Interaction Design programme. You're also a guest professor at Beijing Normal University and a visiting researcher at Newcastle University. You trained as a designer, but you also have a PhD in Instructional Systems Technology, and also a diploma or degree in educational technology. What's really interesting about your path is that you've worked in a variety of roles : Art director, contract designer, trainer, which informed your research on design activity and how design capability is learned. Your research focuses on the way in which the pedagogy and practice of designers informs the development of design ability, and particularly in relation to ethics, design knowledge and learning experience. So we're really thrilled to have you today because you're also obviously an expert in dark patterns. You've consulted on multiple legal cases relating to dark patterns and data protection, and you work with regulatory bodies and nonprofit organizations to increase awareness and Action even more importantly, against deceptive and manipulative design practices. Have I forgotten anything Colin?

‍

Colin M. Gray 

That's a pretty long list. But thank you for the very kind introduction.

‍

Marie Potel-Saville 

First of all, Colin, could you share with us what led you to start working on dark patterns ? 

‍

Colin M. Gray

So my original efforts relating to dark patterns really stemmed from my broader interest in ethics and designer responsibility, some work that I had started during my PhD. And in the last year that I was working on my PhD, I came across a student poster from a master's student, actually here at IU where I did my PhD, Jared Forney, who had done a master's thesis project on dark patterns. And it was the first time I had seen the term, maybe I'd missed something on the internet during my PhD. But I was very struck by the term and by the thesis work that Jared had done. It ended up being a really nice entryway into studying design ethics and technology ethics more broadly. So when I found myself at the start of my assistant professor position, just up the road about two hours at Purdue University where I've been for the last eight years, I wrote a grant on dark patterns really trying to figure out how can we better understand this term, how can we use it as a gateway to study designer ethics more broadly. After a couple of attempts that grant was funded, and I was able to get to work starting in the late spring of 2017, trying to understand dark patterns. And then obviously it led to these broader efforts to understand how we can support designers as they seek to make ethical action in their workplace, and in some cases, even have to understand that they have an ethical responsibility in the first place.

‍

Marie Potel-Saville

Ethical responsibility is everything, isn't it? That's really impressive. Thank you, Colin, for sharing with us. Now, unfortunately, and as you know very well, dark patterns are everywhere. Like some sort of digital plague. You know, the European Commission found that there is at least one dark pattern in 97% of Europeans favorite sites. It's about the same in the US, depending on what counts as a dark pattern but the FTC is really taking a strong stance against dark patterns, which also talks about the importance of the topic. What really struck us when we started working on dark patterns in our r&d lab, was how dark patterns contravene most of the principles of good UX design. We're lucky to have very talented UX designers here at the agency and we actually did a short experiment. You know, when the European Data Protection Board published its draft guidelines on dark patterns in social media, that it was in 2022, to answer the public consultation on these guidelines, we created a user journey for setting up an account on Tiktok, which was an interesting choice, I guess. But anyway, we identified in each of the steps of the user journey the potential dark patterns and we compared them with good UX design principles and also with GDPR. It was really obvious that most dark patterns are actually simply contrary to good UX design principles such as guidance, readability, explicit controls, law of similarity, etc. So Colin, could you share with us how the designer in you explain the huge prevalence of dark patterns around the world whereas they're basically bad UX? What went wrong?

‍

Colin M. Gray

Well, there's a lot that's happened with the internet over the last 15 years. And so I mean, I think there are a number of forces that have led us to this current moment. One of them, which I don't think people tend to talk about too much is the standardisation of the web. So we've seen the rise of design systems which allow these exploitations to occur really at a large scale and they can be shared across many partners across many platforms very quickly. And so I think that is actually one of the first movements in the early 2010s that actually started to lay the groundwork for the prevalence of these dark patterns. The other thing that we're seeing is, I think, a goal for companies to make profit, which of course has always been in place, but as these systems have scaled up, there's been increasing pressure to find ways to monetize platforms of all types. And some of this relates to the concept of growth hacking, where you're trying to just eke out a little bit more profit by making small decisions that are carefully considered through A B testing or other kinds of metrics. And then there’s also some bad actors that have led the way in sort of testing the waters and seeing whether anybody is going to notice. So that's sort of the backdrop against which a lot of these design decisions have been made. Now, in the research work we've done, actually talking to lots of designers and technology practitioners, we found that there are very few that are just blatantly bad actors, there to just hurt the user in any way they can. In many cases, they are actually bringing these things up to their employers or their design teams, these best practices are knowingly being violated and, in many cases, there's just no recourse for them, it's you do this, or we'll find somebody else that will, often people's job is on the line. So people have to make very pragmatic decisions about their own livelihood in absence of regulatory action, which fortunately, is now finally coming.

‍

Marie Potel-Saville 

That's exactly what we've found in our own research with designers. They usually know what's wrong. But they, as you said, they have little margin of manoeuvre. And also, what we found is that they're, they're lacking arguments, readymade arguments, to try to convince decision makers not to enter into deceptive design. So we're also working a lot in this respect to provide designers with some sort of tool boxes with ready made, legally grounded and strong arguments to change the game really. And to your point about making profit, of course, companies should grow and make profits. We had a very interesting conversation in the previous episode with two economists at Oxera Consulting, Anastasia Shchepetova and Lirio Barros, who actually explained that contrary to what most company could think, at first glance, deception by design is actually not profitable in the middle to long term, it might give you a short turnover boost, but very, very short term. But more often than not, the backlash, you know, the customers being furious and cancelling their subscription and flooding your call center, totally overweight the very small advantages and profits that might have been gained at the beginning. So that's a really interesting point in any event, and I'm sure we'll have further discussions on this. Now, you know, specifically talking about UX designers, do you think that there is something missing in UX designers training to adequately address and avoid deceptive design, perhaps to give them more power as well ? Because you said that they have this very little margin of manoeuvre and on a very competitive market, if they don't want to create a dark pattern, probably someone else will. So is there something that can be done in terms of training at all?

‍

Colin M. Gray

So there's certainly something that we can do on the training front to provide designers with more tools to play around with, as you mentioned. Some of it is about lines of argumentation that are successful, some of it is about building robust theories of interaction that allow this kind of ethical inspection of the design work that's going on. One of the bigger challenges I think in the UX space specifically, is that there is no specific line of education that is typical for a UX designer. Now, this has changed a little bit over time, but many UX practitioners, some pick the number over 50%, have no formal training in UX, they've learned it on the job they pivoted from another career. And so they might have pivoted from software engineering, or from marketing, or graphic designers, many other fields, where they may have not ever given a lot of thought to these ethical considerations, they may not even consider it part of their job. There are, you know, mainstream UX programmes I've contributed to that work over the last decade, and trained hundreds of students myself that have gone out into UX practice. And certainly ethics and values are part of the conversation we have with students, but what we found is that knowledge that these things are happening is not enough. In fact, very early on in some of my dark patterns research, we ran design protocol experiments with students, including some of my own students, and we asked them to perform some unethical tasks and it was really interesting. Across multiple types of tasks with multiple framings, with multiple compositions of students, the outcome was always the same. Students didn't feel like they could reject the design brief they were given, they didn't feel like they had the agency and autonomy to reframe the problem. Now, obviously, this is an experimental setting, you know, it's not quite naturalistic, but if students don't feel like they can push back in an educational setting, where can they feel like they can push back? Their job wasn't on the line, their status as a student wasn't on the line, they weren't even getting a grade. And still, you know, they sort of perpetuated these deceptive practices.

‍

Marie Potel-Saville

That is so interesting. Actually, in our training against dark patterns, there's a first exercise that is about creating as many dark patterns as possible. So it is only for training purposes, but we want to see how far they can go in a dystopian scenario and then, of course, we give them the tools and the methodology to avoid creating dark patterns and instead create fair patterns. But it is so interesting, the understanding that they cannot reject a design brief. I guess that talks a lot as well about the place of designers and companies. You know, one could think that in in design lead companies, they would have more agency.

‍

Colin M. Gray

Sometimes they do. And I mean, I think our work has progressed quite substantially beyond that moment that I referred to with the with the lab protocol study, that was done about six years ago. And it really changed some of our pedagogical practices, and the ways that we help students sort of think through and be reckoned with, and what we're increasingly calling wrangling ethical complexity. Because it's something that you have to actively sort of address and try to figure out but you're never going to get rid of the threat, right, the threats are just going to change, change shape and change form. And so some of the techniques that you mentioned, like, you know, intentionally designing dark patterns so you understand what that feels like, are really important. And it actually builds on a well known creativity technique called reverse brainstorming, where you're trying to design the worst possible experience. I think, you know, this needs to be part of our repertoire, these critical and speculative methods that allow designers to very quickly think about not only what is the best experience that I can envision for a user, but also what are some really problematic experiences that I could create, that maybe even sometimes feel like they're legitimate, or they're meeting some goals that are not the user's goals, and really confronting some of those questions around value tensions. And that's, I think, a really important thing, as we also equip practitioners. Because as I mentioned, many UX practitioners, many product designers, and people in sort of similar kinds of roles, were not necessarily trained in that discipline, they might not have that ethical training to begin with. And so we also have the need to equip them. So over the last couple of years, we've really been focusing on building a set of ethics focus methods to really support that not only that awareness, raising for people that need that awareness in the first place, where people who need to share that awareness with their design teams but also giving them ideas about how they can actually put that into action as well, because that's often the critical step that is missing or is not well supported by existing methodologies.

‍

Marie Potel-Saville 

Exactly taking action is the most important. Talking about action, I have to say that the Federal Trade Commission is quite impressive, it is obviously taking a very strong stance against dark patterns. For example, they recently sued Amazon for what is possibly the most dry and dark pattern : Prime. Most of you know what Prime is, it is this subscription model that is heavily pushed as a free trial and free delivery offer, with actually a lot of missing information when you subscribe to it. And then it's extremely difficult to cancel, to the point actually, that the FTC pointed out in its complaint, that Amazon in a very cynical way called prime internally “Iliad” because it's just epic to try to cancel it. So it's not actually funny, because as we all know that kind of dark patterns can cause very severe financial harms. That's precisely why the FTC is ramping up its enforcement. They also imposed heavy penalties for example, on Epic Games. That was a 520 million settlement. ISP, I think it was 100 million penalties. Vonage for 100 million as well and Publishers Clearing House most recently with 5 millions. So all of these figures you know, could seem impressive as such, but what are your views on this enforcement? Do you think that it's working is a deterrent enough? Does it make sense to companies and to designers?

‍

Colin M. Gray 

So I'm very happy to see this enforcement. I think it's a long time and coming. There’re people that I've heard from the Federal Trade Commission that are excited about this sort of new phase of enforcement that really didn't exist five years ago for many political factors in the United States, and sort of the ways that some of these issues were taken up and brought to the courts. I hope that these are examples of many more enforcement actions that we will see in future. These settlements look impressive, and some of them are, I mean, the epic game settlement was quite large. But also, when you look at the financial profile of these companies, they don't seem too bothered either. And so, this is also the case with Google who I was on the opposing counsel team for with the state of Arizona case. In the state of Arizona case plus over 30, other state settlements resulted in Google paying out over $500 million to various states, US states, and hasn't really changed some of their core actions. It doesn't appear so. And many people, including economists, have suspected that large companies in particular, build these kinds of legal remedies, legal sanctions, into their share price and into their stock price in the way that they do business. So, you know, sort of at the high level, a lot of these companies seeing these large sanctions isn't necessarily a sign that everything's going to magically get better. I think we need many other kinds of actions at many different scales, to really impact these these issues that we're sort of reckoning with right now. One thing that's I think, especially interesting about the US context as opposed to the EU context, is that all these settlements are really around consumer protection issues. Very few of them are about data protection issues, is this is different than what we're seeing in the in the EU that's playing out, and it actually shows that there could be a place for much more substantial power by the FTC if they were to go after some of these data protection concerns to the same degree as they're going after consumer protection issues. 

‍

Marie Potel-Saville

Absolutely. To echo your point, I've been a lawyer for a long, long time and I know firsthand that penalty risks are sort of costs of doing business unfortunately for a number of companies. Which is why it's so interesting to think about the other incentives for companies to avoid dark patterns and switch to something else. That could be customer lifetime value, for example, the fact that they are damaging their brand, and that ultimately, they're actually decreasing the value of their company. So that's another conversation but equally interesting. Now, on to a very interesting question, you created with Christiana Santos and Natalia bjelovar, the very first ontology on dark patterns in which you identified three levels of pattern granularity : high level patterns, which are basically general strategies, mezzo level patterns, which describe a given angle of attack, and low level patterns, which include specific means of execution. This was absolutely music to my ears and a great step towards creating this common language among scholars and regulators. Could you share with us why and how you created this ontology?

‍

Colin M. Gray 

So this ontology, actually, our main work started about this time last year. So it's still quite recent work. And it's something that we're still working on. But the rationale for it, I think, stemmed from a number of different things that were happening all at the same time. One was that I was reviewing papers from dark pattern scholars, many which are now available to the public, and they were creating novel dark patterns, which were actually just really repetitions of known dark patterns, but in new contexts. And we were also seeing regulators calling dark patterns by different names than academic scholars who had been studying them for years. And then in other cases, some people didn't agree on what the full set of dark patterns might be, you know, did all dark patterns have to be deceptive, for instance? Or could coercive patterns also fit within sort of the broader umbrella? And so all these conversations were happening at once. What really sealed the deal for me, like this project is what we really need to do, was getting deposed by Google's lawyers in the state of Arizona V Google case last summer. Because in some of those documents that came out, and them trying to reject my testimony and my expert report, one of the big features was essentially them making the claim that nobody can decide on what a dark pattern is. Which is not true. They were taking advantage of this diversity of terminology to say that nobody can agree and so therefore, it's junk science. I think, the big lesson for me is that you probably shouldn't say things like that to a scientist, because then they'll go do the work to prove you wrong. Which is exactly what we did. So we spent last fall looking at a number of regulatory reports that then came out. We looked at five different regulatory reports that included taxonomies, many of them, you've mentioned already, and five of the key academic texts that had also been referenced in those taxonomies, either directly or indirectly. We used that to sort of create a coherent set of documents and pattern types to start with, and then we went through and we organise them. It was long and painstaking work to look at the definitions, to look at the names and figure out where there were overlaps, where things were being called something different but were actually the same thing. And then I think the biggest thing that came out of that work was recognising that there is some hierarchy to these pattern types as well. I had, of course, proposed this back in my 2018 paper on dark patterns, which is broadly cited in other studies. Where I'd propose that there were these sort of general level designer strategies like sneaking force, action, interface, interference, and so on, but then there were also these low level dark patterns, which were examples of those larger designer strategies. So we really just expanded on that kind of hierarchy and we've realised about halfway through the project, there's actually not just a high and a low level, there's a mezzo level too. And so what we realised through the mezzo level creation was that this was an opportunity not only to categorise things a little bit better, but also to future proof the ontology as well. And so the way that we really went about it is that many of the new studies that are coming out on dark patterns in new contexts, with new technologies, with new user groups, were really repeating some of the same known dark pattern types, but at a low level. So this is the execution level that happens in code, or in the user interface or what a user sees. But we were able to abstract that and identify what is the mezzo level pattern or the angle of attack that's happening here. And so you know, for instance, within sneaking, a lot of people talk about Harry's original pattern : sneaking the basket, or something appears in your shopping basket at checkout that you didn't put in there. And so of course, that's part of the broader high level pattern of sneaking, but there's an important mezzo level, which is that information is being hidden from the user. And so as soon as we've identified that mezzo level pattern, we can use our imagination to think about many other ways that designers on their bad days, find many other ways of hiding information from consumers. And so those can be used to articulate new low level patterns that are bound up in specific contexts, or specific sort of use or domain areas as well. So we're just getting started. 

‍

Marie Potel-Saville 

You know I am a huge fan of your work. I said that many times. And indeed, future proofing the ontology is extremely, extremely important. What are the further steps planned for the ontology? 

‍

Colin M. Gray

We're really trying to build a community around the ontology work, hopefully, we'll be launching a website with the ontology materials later on this year. Right now, we are rapidly writing because of the CHI deadline. That's the large human computer interaction conference where we published the ontology at last year. The paper deadline is coming up in just a few short weeks and we'll be planning to submit a longer form version of the ontology. The first steps that we're taking, are, again, to build this larger community and allow people to really use the ontology. In the work that got published earlier this year, we provided the classification and the linking back to the original sources but we weren't able to reconcile all the definitions. So what we'll be providing in this more advanced version of the ontology is : grounded structured definitions for each low level, mezzo level and high level pattern, and we'll also be providing a sentence template so that all definitions are consistent and have all the same elements. So they can be really easily compared against each other and that also they can be used to support other scholarly and regulatory work. The other piece that we'll be doing with this expansion of the ontology is providing some examples of how scholars who are identifying dark patterns and new work, which we celebrate and try to support them in the best ways that we can, can use the ontology to figure out what is already well known what is well established? Is this just a further piece of evidence? And what is legitimately a new dark pattern and how might it be added to the ontology?  So that we can continue to extend in particular, the low level pattern examples that we have to work with. 

‍

Marie Potel-Saville 

Brilliant and I'm sure it will be extremely valuable for regulators and also perhaps in-house lawyers, if they're smart, so that they can better advise their client. When we worked on trying to identify countermeasures to dark patterns, we had the exact same issue. We realised there were so many taxonomies and so this need for a common language is absolutely clear and obvious. What about the countermeasures? What do you think could bring sustainable change on dark patterns?

‍

Colin M. Gray

You're asking me the complicated questions this morning, but I'll do my best. Now, this is a really important thing. Not only we need to be able to push back and say no, that's not ethical, that's problematic, but we also need to be able to present design alternatives that still bring value to organisations. As long as we live within a capitalistic world, which we do, we need to equip designers with more tools than just rejecting, but also providing new ideas and putting them out in the world. And so the ontology, I think, is a really nice point of departure to link these countermeasures that I know that you and your team have been working on. Many of these countermeasures that I've seen already, focused primarily on the user interface layer, because that's what's most easily adaptable and changeable. And so many of those are going to link with the high level strategy of interface interference. What I'd love to see in in coming years is really more detailed inspections, sort of at the product management level, and the strategy level on how we can effectively fight back against some of the other high level pattern types as well. And so when you look at, particularly the high level patterns of : obstruction, sneaking and forced action, all of those are temporarily bound. And often they're out of the control or even out of the eyesight of users. And this is something where you really do need to approach it from the strategy level. So providing just an alternative pattern isn't really going to work in those cases, it might actually require an inspection of an entire business model. Then sort of similarly, with nagging. This sort of speaks to the attention economy and these broader issues and concerns about what it means that we're distracting ourselves with technology in the ways that we are. And I think this is going to come under increased scrutiny and might also lead to some broader action that's needed to figure out under what conditions is it okay to sort of play with people's time as a resource and not just focus on the economics of the situation. 

‍

Marie Potel-Saville

Absolutely. And to your point, you know like the OECD is doing, we could even ask ourselves : Is it really a sustainable digital economy to trick and deceive and manipulate people, even to make decisions ? It's not just about losing that time, it's also to make decisions against their own interests. I'm sure you know, but the European Commission found that some dark patterns actually create severe health harms, like raising the heart rate, raising anxiety, you know, they ran IRM and scans, that really shows that there's a particular part of the brain that goes all red, when users are subjected to dark patterns. So really, it's a matter of sustainability. I fully agree with you. How might we create a sustainable digital economy where, of course companies can grow and make profits, but not to the detriment of humans?  But on the contrary, in an environment where humans can actually thrive? That's the key topic. So that's a quite a high level conversation but back to a very practical and final question, which is now traditional in this podcast : Could you share with our audience, some very practical tips for designers to avoid creating dark patterns or remedy existing ones?

‍

Colin M. Gray 

Absolutely, I get the privilege of training designers every day. And so this is a conversation that we have often. So I think some of these things have already been mentioned throughout the interview so I'll sort of package them back together, maybe add a couple of additional ones. One piece is actually just having the awareness that these dark patterns and these practices exist in the first place. And recognizing that these are ethical tradeoffs that are being made by a designer when they choose one design pattern over another. Really understanding like, what are the broader impacts on the user on society when you're undertaking some of these key design decisions? A second is, understanding where design goes wrong as well. This is where the speculative or critical methods can really be useful in exploring the design space and not limiting yourself trying to be perfect but also seeing where things go wrong. You learn a lot through those contrasts that might help you train your brain a little bit better to realise when you're sort of stepping over an ethical boundary that you just might not have been sensitised to before. And then I would say the final thing is, once you have some of those other bits in place, you know what's going on, you know that their ethical questions, it's really on you to educate your team and your company about the importance of these issues as well. This is, I think, where we need more resources, not only to get a design team on the same page, and make sure that everyone regardless of disciplinary background, recognizes that these are ethical questions, but also to use many forms of argumentation that may work differently in different organizations or in different countries. So knowing when it's probably a good pattern of argumentation, to maybe say, this might put us at legal liability if we do this, here's an example of a case from deceptive dot design that's been pulled together to show you know what some of these sanctions look like. Or, you know, maybe some evidence about these economic models to show this may work in the short term, but it's actually going to a lead to, you know, long term issues with customer retention. So it's really a designer thinking about that full gamut of opportunities to interact, to educate, to lead, and to hopefully, inform the social responsibility of the products they create. 

‍

Marie Potel-Saville 

Look, I absolutely love everything you said and in particular the tools to be given to the designers. We happen to be creating this toolbox of arguments, depending on the audience, because we've seen in our research exactly as you did that, depending on who designers are talking to, they need different lines of argumentation. So we're precisely building the toolbox. Bear with us. It will be released soon. You'll be the first to know. Thanks a million Colin, it's been my absolute pleasure to have you today. Thanks again. And yeah, let's definitely stay in touch.

‍

Colin M. Gray 

Sounds great. Thank you so much for having me on today.

‍

Marie Potel-Saville

Well, thanks a lot for listening today and stay tuned for our next episode.

‍

Fighting dark patterns, a podcast by Amurabi. For further information, you can go to fairpatterns.com

‍