The tech sector has a diversity problem. For a long time, the overwhelming whiteness and maleness of Silicon Valley’s workforce was just accepted as normal — if it was acknowledged at all. Some even argued this was the result of meritocratic decisions where the best man for the job was always chosen. Now those views are, at last, being recognised as products of social inequities and rejected by tech companies vowing to “do better” on the issue of diversity.
Yet “doing better” shouldn’t just involve hiring more gender- and racially diverse people. The diversity agenda must also broaden to include progressive ideas and incorporate them as core features in the design of technologies we use every day. There are obvious examples where this is already on Big Tech’s radar, such as removing algorithmically-generated gender or racial bias in job ads, or changing the way digital voice assistants respond to sexual harassment. But what about issues that are more subtle and pervasive: like how technology design deals with user consent?
The issue of consent provides an important example of how we can rethink everyday technological interactions by replacing outdated approaches with progressive practices propelled into the mainstream by movements such as #MeToo. The way technology consent is designed now — such as when we click “agree” on a software licence or “accept” for data collection — is more dismissive than anything. Consent has become just a box that has to be checked. It functions more like agreeing to acquiesce, rather than actually respecting people’s autonomy.
This model of consent contradicts essentially every principle of affirmative sexual consent for which feminist groups have long advocated.
In new research which has recently been published, we argue that the tech industry needs to go beyond consent events that focus on obtaining one-time, “yes” or “no” consent at the start of a technological interaction, such as during product sign-up. As more of the things we use every day become more intelligent and active, it is more critical than ever that we throw out retrograde models of consent and implement progressive alternatives. Consent must be understood as an ongoing and voluntary process of negotiation, which is regularly reaffirmed and where opportunities for exercising autonomy are always available.
In our article, we propose an alternative model inspired by sexual consent best practices from the BDSM community — bondage, domination, submission, masochism — which has been at the forefront of not only understanding but implementing consent as an open dialogue, rather than a single decision. To be clear, the BDSM community is not a perfectly progressive place. It has confronted a lot of criticisms and problems due to exploitative power dynamics and violence in some of its activities. With that recognition, there are valuable lessons to learn how to improve the status quo of consent.
We offer a series of best practices from the BDSM community — soft/hard limits, safewords, traffic lights, and aftercare — and explain how they can be applied to make consentful tech by rethinking and redesigning human interactions with internet-connected things like smart fridges and voice assistants.
First, there is the setting of soft and hard limits. In BDSM practices, these are parameters for which activities are allowed or prohibited during a particular sexual encounter. In technology, the ability to set limits empowers people to take ownership over the terms of an interaction by determining and redefining their own parameters. This is especially important because many smart things are now designed to “nudge” us toward acting in certain ways. Some data collection, persuasion, and manipulation may be okay within reason, whereas others might not be at all. For example, perhaps somebody has no problem with advertisers accessing the data collected by their smart fridge for the purpose of providing targeted coupons based on consumption habits, but it might cross the line for an insurance company to use the same data to incentivise behavioural changes or adjust premiums. For people to be able to set these limits in the first place, companies also need to be transparent and explicit about what they are doing or want to do with us. Duplicity is not conducive to consent.
Second, there is the creation “safewords”. In sex, safewords are a method for immediately withdrawing consent and stopping all activities — regardless of already negotiated agreements — by saying an out-of-context word or phrase like “Vanilla”. In technology, safewords can be a quick way of turning off technologies that are designed to be always on by default. Just as devices like Amazon’s Alexa have “wake words” that activate their voice command capabilities, so too should smart technologies have safewords that make them stop collecting and sharing data, or turn off their always-listening mode, until we decide to turn them back on. Effective safewords should stand out from normal commands or requests so they are unlikely to be misunderstood. Something like: “Alexa, aardvark”. Safewords allow people to take control of voyeuristic devices by ensuring they stop watching us when we want them to.
Third, there is the establishment of traffic lights. These recognise that consent is a spectrum with considerable middle-ground that is likely to change depending on context. This design feature is meant to emphasise the ability of users to be able to exercise their autonomy at any time by changing how technology interacts with them. A traffic light system would provide simple, ongoing options for users to indicate where they want an interaction to continue (green), slow down or “check in” with them (orange), or stop (red). Such options may already be buried in the privacy settings within the appliance’s software. But they should be readily apparent and easily accessible, rather than making people hunt through multiple menus just to make choices about agency and consent.
Finally, there is engaging in aftercare, which recognises that interactions with technologies elicit various emotional responses that may leave their users in need of additional support. Interacting with a digital voice assistant may, for instance, raise concerns and questions for users they want to discuss and resolve. A strange encounter, or a device malfunction, may prompt concern (like when Alexa started randomly laughing in some users’ homes). Likewise, forming intimate relationships with voice assistants — as some people are already doing and as some companies are promoting — may leave people emotionally vulnerable. Having a process for aftercare could involve interactions or check-ins with a voice assistant like Alexa itself, referrals to other types of help, or an opportunity to provide feedback to the device’s manufacturer.
Rethinking critical issues like technological consent doesn’t mean starting from scratch; nor does it have to mean burdensome and laborious consent interactions. We need to engage with and learn from other communities who have long been advocating for better alternatives. By applying feminist ideas to design, our everyday interactions with technologies can be made into more empowering and equitable experiences. The tech sector doesn’t only need more diversity in terms of people; it also needs to take more seriously the diverse range of ideas and issues that have traditionally been considered “women’s concerns”.
Yolande Strengers is a digital sociologist and human-computer interaction scholar investigating the sustainability and gender effects of digital, emerging, and smart technologies. She is Associate Professor in the Department of Human Centred Computing at Monash University, where she leads the Energy Futures research program. She is co-author (with Jenny Kennedy) of The Smart Wife: Why Siri, Alexa, and Other Smart Home Devices Need a Feminist Reboot.
Jathan Sadowski is a Research Fellow in the Emerging Technologies Research Lab at Monash University. His work focuses on the political economy of smart technologies. He is the author of Too Smart: How Digital Capitalism is Extracting Data, Controlling Our Lives, and Taking Over the World.