The KT4D project defined its impact as flowing from two key outcome-focussed pathways:
● Outcome/Impact Pathway 1: ‘Protect fundamental rights and European values from possible threats stemming from unregulated use of artificial intelligence (AI) and big data applications’;
The KT4D contributions to the protection of rights and values from possible threats stemming from unregulated use of artificial intelligence (AI) and big data applications commences from the fundamental work produced by WPs 3 and 4. The second and final releases of the first six modules of the Social Risk Toolkit (KER 1) have been achieved in this reporting period, sharpening and expanding our consolidated, user-friendly baseline from which to re-imagine approaches to the use and regulation of big data and AI in democracy-adjacent contexts. In particular, these WPs shifted from a mode of assembling a baseline of existing research to a mode of filling in the gaps in the interdisciplinary knowledge base related to AI, big data and democracy. This was further enhanced with the targeted inputs from WP6, which addressed specific issues of interest at the intersection of culture and technology, in specific questions of critical digital literacy and the role of local cultural contexts in both technology and democracy.
This background and targeted empirical and advanced research were harnessed to deeply inform the Governance framework, policy roadmap and recommendations produced by WP5 (KER 2). This powerful suite of policy-focussed tools lays out a clear vision for AI as supporting democratic society by 2035. The actions proposed are clear, bounded, granular, and categorised by timeframe and type. The proactivity that informs and enlivens these assets have been co-developed with and warmly welcomed by end-users, to support imaginaries and actions beyond the harm-reduction measures of GDPR to foster democracy promotion via value-based frameworks in platform technologies.
This impact pathway was also supported by the conceptual development and learning materials created by WP6 (KER 4). These results take the subtle but well-formulated concept of critical digital literacy and realise it through a range of assets able to support different vectors for the development of both understanding of technology in democratic contexts as well as individual and collective agency in this context. It draws on a clearly defined understanding of threats, proposing and building mechanisms by which citizens (and organisations that represent and provide technology upskilling opportunities to them) can recognise and resist them. This work has also been tested and validated with user communities in the course of this reporting period with two groups of citizens from two different national/linguistic contexts (in Madrid and Warsaw) through KT4D Use Cases 2 and 3. This close interaction with user groups has been a key component in KT4D’s strategy to develop impact.
● Outcome/Impact Pathway 2: ‘Innovative uses of AI and big data to enhance citizen engagement and democracy.’
The Social Risk Toolkit drew together insight regarding the granularity at which advanced knowledge technologies and democracy interact. Through this process, a number of key issues for the adoption of useful frameworks were uncovered: in particular, the fragmentation of the concept of literacies proved an important point to clarify, as critical digital literacy and digital literacy often become conflated not only with each other but with terms such as information or media literacy.
The DDL makes a specific contribution to the mission of piloting innovative uses of AI and seeding new approaches to it. Unlike many other projects in this space, KT4D did not focus so much on creating a platform we could then license for use. Instead, we focussed on democratic purpose and how this would interact with technology. As a result, the project created a powerful ‘tool for thinking,’ in particular about how software quality and fitness for purpose might be judged (overthrowing, for example, the privileging of efficiency as a goal for software function), making a significant contribution to how HCI in a collective sense might be conceptualised in this sensitive context.
The KER 3 interactive digital narrative achieved a similar goal of expanding conceptualisations of how a ‘good’ software tool would function, and what considerations might be taken into account in this respect. The relevance of this work was validated through all of the use case interactions with the DDL, and the final valourisation meeting of UC4. It encouraged a deeper level of self-reflection about the integration of ethical practices into the work of software developers and policymakers, in particular in the light of the many tools that exist to potentially guide this relationship, most of which are not only not used, but not known to the community. The UC4 meetings also yielded some further surprising results regarding the kinds of tools that software developers find most useful, and the kinds of cultural issues they feel least equipped to deal with.
