Bookmarked AI Liability Directive (PDF) (by the European Commission)

This should be interesting to compare with the proposed AI Regulation. The AI Regulation specifies under which conditions an AI product or service, or the output of one will or won’t be admitted on the EU market (literally a CE mark for AI), based on a risk assessment that includes risks to civic rights, equipment safety and critical infrastructure. That text is still very much under negotiation, but is a building block of the combined EU digital and data strategies. The EU in parallel is modernising their product liability rules, and now include damage caused by AI related products and services within that scope. Both are listed on the EC’s page concerning AI so some integration is to be expected. Is the proposal already anticipating parts of the AI Regulation, or does it try to replicate some of it within this Directive? Is it fully aligned with the proposed AI Regulation or are there surprising contrasts? As this proposal is a Directive (which needs to be translated into national law in each Member State), and the AI Regulation becomes law without such national translation that too is a dynamic of interest, in the sense that this Directive builds on existing national legal instruments. There was a consultation on this Directive late 2021. Which DG created this proposal, DG Just?

The problems this proposal aims to address, in particular legal uncertainty and legal fragmentation, hinder the development of the internal market and thus amount to significant obstacles to cross-border trade in AI-enabled products and services.

The proposal addresses obstacles stemming from the fact that businesses that want to produce, disseminate and operate AI-enabled products and services across borders are uncertain whether and how existing liability regimes apply to damage caused by AI. … In a cross-border context, the law applicable to a non-contractual liability arising out of a tort or delict is by default the law of the country in which the damage occurs. For these businesses, it is essential to know the relevant liability risks and to be able to insure themselves against them.

In addition, there are concrete signs that a number of Member States are considering unilateral legislative measures to address the specific challenges posed by AI with respect to liability. … Given the large divergence between Member States’ existing civil liability rules, it is likely that any national AI-specific measure on liability would follow existing different national approaches and therefore increase fragmentation. Therefore, adaptations of liability rules taken on a purely national basis would increase the barriers to the rollout of AI-enabled products and services across the internal market and contribute further to fragmentation.

European Commission

Public Spaces is an effort to reshape the internet experience towards a much larger emphasis on, well, public spaces. Currently most online public debate is taking place in silos provided by monopolistic corporations, where public values will always be trumped by value extraction regardless of externalised costs to communities, ethics, and society. Today the Public Spaces 2022 conference took place. I watched the 2021 edition online, but this time decided to be in the room. This to have time to interact with other participants and see who sees itself as part of this effort. Public Spaces is supported by 50 or so organisations, one of which I’m a board member of. Despite that nominal involvement I am still somewhat unclear about what the purpose of Public Spaces as a movement, not as an intention, is. This first day of the 2-day conference didn’t make that clearer to me, but the actual sessions and conversations were definitely worthwile to me.

Some first observations that I jotted down on the way home, below the photo taken just before the start of the conference.

a) In the audience and on stage there were some known faces, but mostly people unknown to me. Good thing, as it demonstrates how many new entrants into these discussions there are. At the same time there was also a notable absence of faces, e.g. from the organisations part of the Public Spaces effort. Maybe it’s because they rather show up tomorrow when the deputy minister is also present. As an awareness raising exercise, despite this still being a rather niche and like minded audience, this conference is certainly valuable.

b) That value was I think mostly expressed by the attention given to explaining some of the newly agreed European laws, the Digital Markets Act, Digitals Services Act and Data Governance Act. For most of the audience this looks like the first actual encounter with what those laws say, and one panel moderator upon hearing its contents showed themselves surprised this was already decided regulation and not stuck somewhere in a long and slow pipeline of debate and lobbying.

c) It was very good to hear people on stage actually speaking enthusiastically about the things these new laws deliver, despite being cautious about the pace of implementation and when we’ll see the actual impact of these rules. Lotje Beek of Bits of Freedom was enthusiastic about the Digital Services Act and I applaud the work BoF has done in the past years on this. (Disclosure: I’m on the board of an NGO that joins forces with BoF and Waag, organiser of this conference, in the so-called Digital Four, which lobbies the Dutch government on digital affairs.)
Similarly Kim van Sparrentak, MEP for the Greens, talked with energy about the Digital Markets Act. This was very important I think, and helps impress on the audience to engage with these new laws and the tools they provide.

d) The opening talk by Miriam Rasch I enjoyed a lot. Her earlier book Friction E felt seemingly lacked some deeper understanding of the technologies involved to build the conclusions and arguments on, so I was interested in hearing her talk in person. The focus today was more on her second book Autonomy. I‘ll buy bought it and will read it, also to clarify whether some of the things I think I heard are my misunderstanding or parts of the ideas expressed in the book. Rasch positions autonomy as the key thing to guard and strengthen. She doesn’t mean autonomy in the sense of being fully disconnected from everyone else in your decisions, but in a more interdependent way. To make your own choices, within the network of relationships around you. Also as an emotionally rooted thing, which I thought is a useful insight. She does position it as something exclusively individual. At the same time it seems she equates autonomy with agency, and I think agency does not merely reside on the individual level but also in groups of relationships in a given context (I call it networked agency). It seemed a very westernised individualistic viewpoint, that I think sets you up for less autonomy because it pits you individually against the much bigger systems and structures that erode your autonomy, dumping you in a very assymetric power struggle. A second thing that stood out to me is how she expresses the me-against-the-system issue as one of autonomy versus automation. It’s a nice alliteration, but I don’t accept that juxtaposition. It’s definitely the case that automation is frequently used to dehumanise lots of decisions, and thus eroding the autonomy of those being decided about. But to me it’s not inherent in automation. When you have the logic of (corporate) bureaucracies doing the automation, you’ll end up with automation that mimics that logic. If I do the automation it will mimic my logic. I use automation a lot for my own purposes (personal software), and it increases my agency, it’s a direct expression of my autonomy (or that of the groups I’m part of). There’s more to be said, in a separate post, a.o. about the 3 or 4 thinking exercises she took us through to explore autonomy as a concept for ourselves. After all it wouldn’t make us more autonomous if she would prescribe us her definition of autonomy, precisely because she underscores that it’s not a purely rational concept but an emotional one as well.

e) Prof. Tamar Sharon of Radboud University spoke about the influence tech companies have in other domains than tech itself because of their technology being expanded into or used in those domains such as health, education, spatial planning, media. She calls it sphere transgressions. This may bring value, but may also be problematic. She showed a very cool tool that visualises how various tech companies are influential in domains you don’t immediately associate them with. A good thinking aid I think also in the upcoming discussion about sectoral European data spaces and being alert to the pitfall of it turning into a tech dominated discussion, rather than a societal benefit and impact discussion.

f) Kudos to the conference organisers. Every panel composition was nicely balanced, it shows good care in curating the program and having tapped into a high quality network. I know from experience that it takes deliberate effort to make it so. Also the catering was fully vegetarian and vegan, no words wasted on it, just by default. That’s the way to go.

In the Open Data arena people often ask if ‘the people’ are actually ‘ready’ to deal with the availability of data. Do we have the statistical skills, the coding skills, to make data useful?
In my presentations over the past 8 months I’ve positioned data as an object of sociality: it becomes the trigger for interaction, a trigger for the forming of connections between people. Much like photos are the social object of a site like, and videos are the social object of YouTube, or your daily activities are for Twitter.
The current best example of how data can be a social object is something John Sheridan showed at the Vienna Open Data Conference last June. All legislation information in the UK has been made available as linked open data. This makes it possible to reference specific paragraphs in laws.
In general law is generally regarded as boring and decidedly un-hip, but the availability of all this legal data as linked open data has a surprising effect: people are referencing specific paragraphs in their on-line conversations, for instance on Twitter. This is what you see in the screenshot below, where people link to specific parts of UK legal texts in the course of their conversation. From boring and useless texts (other than to legal minds that is), to the social object around which everyday conversation can revolve.

Data is a social object. It is a trigger for citizen participation that way, a new way for people to engage with their community. And, the other way around, participation (e.g. existing participatory processes, existing conversations) is a path to data use. From this basic starting point any newly needed skills will grow.