Oaths, pledges and manifestos: a master list of ethical tech values
Digital technologies are everywhere: in our lives, our homes, our work, our infrastructure. They move fast, and they’ve never been easier to deploy: “We’re not hand-crafting dovetail joints here,” as Anne Currie wrote in The Register.
Because tech has such potential for impact at scale (and because trust in tech companies is falling), there’s been a lot of conversation about ethics and ethical codes lately. So over the last year, Doteveryone has been researching ethical initiatives in tech as part of our work to motivate the tech industry to be more responsible and accountable to society. Many of these take the form of principles, manifestos, or pledges, setting out what good technology or technology practice looks like.
We’ve written up an overview of what we’ve found, along with commentary on some of our favourites. But before we get into the specifics, we should probably talk about oaths and principles generally.
So you’ve got an oath…now what?
Doteveryone cares about making practical change. That’s why we believe it’s not enough to just create ideals — you need to make them real.
Firstly, enough developers, designers, managers and others in tech need to commit to an oath for it to work. And for those that commit and then break it, especially intentionally, there need to be consequences — some sort of accountability.
That might be the full professionalisation of individuals, as we see in other engineering disciplines and which some internet technology people follow too (IET, BCS, IEEE, ACM). It might be social pressure of some sort. Or it might be something else we’ve not thought of yet — if you’ve got an idea, share it in the comments.
There also needs to be a way for people to find what’s right for them, and have the confidence to know they’ve found it. Lots of people are coming up with pledges and principles, but they vary widely. Some are top-down, some are grassroots-led; some are old, some are new. Some are personal (these tend to be oaths), some are for organisations or teams or are general principles. Some have significant uptake, and some do not. Some include lawfulness, and some don’t (or think lawfulness is implicit). Some are European and some are American. In short, there are a lot.
In our own work, we’ve started to frame what we think responsible tech looks like.
We currently say:
– does not knowingly deepen existing vulnerabilities and inequalities, or create new ones
– protects existing democratic and human rights
– is made by teams that are mindful of their ethical, social and human impact
– has controls in place to react to and guard against unintended consequences
– is designed with security and safety in mind
This obviously needs translation into more actionable forms (focussed on activities or outputs), but here our aim is to set out a high level goal. We also want this definition or principles to work for a wide range of digital, internet and computer technologies, both today and tomorrow.
An oath is one potential way we could translate our ideas about responsible tech into action. It’s also been a popular way for people to capture what good practice looks like when thinking about ethics, or exploring ethics in new domains. So it’s been interesting to look at the elements others have included in their oaths and principles in areas related to responsible technology.
After our review, we’re still pretty pleased with this as a starting point, although from the other oaths and more (see below) there are many interesting ideas we might want to include in more specific tools and principles of responsible tech.
We’re still developing those more specific tools and principles, and we’ll share more about that later. For now, here are some of the most interesting codes we’ve found so far. (Found one we’re missing? Let us know at [email protected].)
The master list
2018 ACM Code of Ethics — Revision
The code draft linked here is under development and review. Unlike many informal citations of the ‘Hippocratic oath’ which tend to assume it includes the statement “do no harm” (which modern implementations usually do not — much medical treatment does some harm as well as benefit), here we see “avoid harm.” There is also mention of legacy systems (a real challenge in technology) and a rich consideration of discrimination and societal consequences.
Adam Greenfield’s “Some ethical guidelines for user experience in ubiquitous-computing settings”
The everyware principles are unusual in that they mention the importance of default options in technology, and the risk of loss of face of users as something to consider. They also use “self-disclosure” as a way of framing intentional transparency for connected devices and environments.
This specifically calls out the importance of thinking about the wider context of one’s work.
Asilomar AI Principles
Whilst these were written to be AI specific, much of the content can be seen as applying to digital technologies more broadly. The inclusion of requirements for investment and the relationships between industry and research are unusual.
Principles for Digital Development
Note that this means international development, not software!
Ethical Design Manifesto
This manifesto requires some quite specific technical features — such as distributed, open source, end to end encrypted. These perhaps go beyond either the level of specificity, and/or set a higher standard, than we are aiming for today with responsible technology.
The Manifesto for Data Practices
There is a practical and design focus here, including making work which can be built upon by others in various ways. However, it’s not clear that followers of these practices would handle personal data in a way which European culture might wish to see — there is an implicit sense that data collection, rather than minimisation, is good.
This was created in response to specific political developments in the USA — as well as talking about design and participation in activities, it includes mention of destroying and scaling back existing technology too
The Obligation of the Engineer
This oath includes a recognition of the power of engineering in society.
EPSRC/AHRC Principles of Ethical Robotics
Although this is about robots, the principles here apply to code or bots in general.
The Programmer’s Oath
This is interesting for the inclusion of a recognition that other domains are challenging too.
Royal Academy of Engineering’s Statement of Ethical Principles
This considers future generations and risks, and also talks about objectivity and truthfulness in statements made professionally.
It is also used by the Engineering Council, and is enforced by UK professional engineering bodies on their registered members, which include tech practitioners (such as the author, who is a chartered engineer with the IET).
This has a real feel and brand to it, both framing the issues and the response of the practitioner and their products to it.
What are we missing? Let us know at [email protected] and we’ll add it to the list.