Data School (Utrecht University)

This recent post on subjecting AI to “job interviews” fits well within Data School's view on the importance of professionalizing the entire “hiring” process, including performing background checks and periodic performance appraisals. Check out our full response here: www.linkedin.com/pulse/job-in...

Job Interviews for AI? What It Really Takes to 'Hire' AI (Response to https://www.oneusefulthing.

The "Job interview for AI" metaphor is useful @emollick.bsky.social but don't underestimate the importance of professionalizing the entire "hiring" process, including performing background checks and periodic performance appraisals. Read our full response here: www.linkedin.com/pulse/job-in...

#ResponsibleAI #AIEthics #AIAct #AIAudit #DataSchool

💬 P.S: For readers working in the Dutch context: we are looking for organizations who would like to participate in a one-year pilot (starting early 2026) to test our AI performance appraisal instrument in practice.
📩 Interested? Contact us at dataschool@uu.nl.

AI auditing through performance appraisals: A practice-informed approach - AI & SOCIETY AI auditing is projected to be a crucial governance mechanism for ensuring fair and responsible use of AI systems in society. Yet, despite a proliferation on the market of well-intended new AI auditin...

👉 𝗥𝗲𝗮𝗱 𝘁𝗵𝗲 𝗽𝘂𝗯𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗵𝗲𝗿𝗲:
link.springer.com/article/10.1...

💡𝗪𝗵𝘆 𝗿𝗲𝗮𝗱 𝗶𝘁?
• Discover how a performance appraisal instrument may help your organization prepare for AI auditing.
• Join the conversation on utilizing performance appraisals for AI systems to comply with the AI Act’s documentation requirements.

💡𝗪𝗵𝘆 𝗿𝗲𝗮𝗱 𝗶𝘁?
• Critically engage with the proposal to develop performance appraisal instruments for AI systems.
• Learn more about one such assessment instrument for AI, developed by Data School, in co-creation with supervisory authorities RDI and NVWA.

The core idea: 𝘦𝘹 𝘢𝘯𝘵𝘦 assessments like the Fundamental Rights and Algorithms Impact Assessment (FRAIA) need to be supplemented with relatively light-weight iterative assessments in which the AI system’s functioning is periodically appraised in relation to the broader organizational context.

In their latest publication, published in AI & Society, Bart A. Kamphorst, Iris Muis, Julia Straatman, and Mirko Tobias Schäfer (@mtschaefer.bsky.social) argue that there is a need in practice for robust, periodic documentation procedures around AI systems – akin to maintaining personnel files.

Now that your AI system is running in production, how do you assess whether its functioning continues to align with your organization’s goals and values? And how do you ensure that the documentation around the AI system remains up-to-date and auditable?

📣 𝗣𝘂𝗯𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗔𝗻𝗻𝗼𝘂𝗻𝗰𝗲𝗺𝗲𝗻𝘁! “𝗔𝗜 𝗮𝘂𝗱𝗶𝘁𝗶𝗻𝗴 𝘁𝗵𝗿𝗼𝘂𝗴𝗵 𝗽𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗮𝗽𝗽𝗿𝗮𝗶𝘀𝗮𝗹𝘀: 𝗔 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗲-𝗶𝗻𝗳𝗼𝗿𝗺𝗲𝗱 𝗮𝗽𝗽𝗿𝗼𝗮𝗰𝗵”

And as these five voices remind us, this ethos ripples outwards and is carried into new roles, new institutions, and new futures.

👉 Read the full blog post on our website: dataschool.nl/en/alumni/

For us, this reaffirms what Data School has always stood for: not only investigating but shaping the digital society, together with stakeholders, across sectors, through hands-on engagement and mutual knowledge transfer that bridges academia and society.

🤝 Learning through actively working with, within, and alongside public sector organisations and being agents of change
✨ Finding clarity in their professional identities and the futures they want to pursue

💡 Being “thrown in at the deep end” and encouraged to take that crucial first leap of faith
🚀 Stepping outside comfort zones, yet having the freedom to explore, experiment, and learn by doing

Five former team members shared what stayed with them from their time at Data School:

🎓 PhDs, PostDocs, and lecturers in higher education and research
🏛 Policy advisors, data analysts, and digitalisation consultants in the public sector and government
📊 Advisors, marketeers, and social media specialists
🎨 Project managers in music & design, café owner, and other independent pursuits

🌱 𝗖𝗮𝗿𝗲𝗲𝗿𝘀 & 𝗟𝗶𝘃𝗲𝘀 𝗔𝗳𝘁𝗲𝗿 𝗗𝗮𝘁𝗮 𝗦𝗰𝗵𝗼𝗼𝗹: 𝗛𝗼𝘄 𝗘𝘁𝗵𝗼𝘀, 𝗣𝗿𝗮𝘅𝗶𝘀, 𝗮𝗻𝗱 𝗩𝗮𝗹𝘂𝗲𝘀 𝗥𝗶𝗽𝗽𝗹𝗲 𝗢𝘂𝘁𝘄𝗮𝗿𝗱𝘀

As we traced the trajectories of 44 former colleagues, we found them today spread across a wide range of organisations, working as:

👉 Direct aan de slag? DEDA is vrij beschikbaar via: deda.dataschool.nl

🚀 𝗜𝗻𝗻𝗼𝘃𝗮𝘁𝗶𝗲𝗳 𝗔𝗜-𝗽𝗿𝗼𝗷𝗲𝗰𝘁 𝗶𝗻 𝗱𝗲 𝘀𝘁𝗮𝗿𝘁𝗯𝗹𝗼𝗸𝗸𝗲𝗻?
Verantwoorde AI keuzes wachten niet — begin vandaag nog met De Ethische Data Assistent (DEDA).

🤝 Extra ondersteuning nodig of een ervaren moderator inschakelen?
Neem contact op: dataschool@uu.nl

#DEDA #Data #Ethics #ResponsibleAI #DataSchool

It's the opening of the academic year in the Netherlands today.

Public trust in science is stable and relatively high 💪

It is higher than trust in any other institution (judiciary, media, business or politics).

Trust in the government is the lowest.

Source: Rathenau in @volkskrant.nl

There is no such thing as a digital twin: Deconstructing sociotechnical imaginaries of digital twin technologies in Dutch urban governance - Sofie de Wilde de Ligny, Sharon van Geldere, Mirko Tobias S... Within the context of three-dimensional simulations of urban space and their variables within, digital twin technologies (DTs) are touted as making urban govern...

So happy to see this paper out on deconstructing sociotechnical imaginaries of digital twin technologies with Sofie de Wilde de Ligny, @mtschaefer.bsky.social & @albertmeijer.bsky.social! 🚀

journals.sagepub.com/doi/full/10....

I might be late to the party, given the citation count, but:
I really enjoyed this exceptional piece on deplatformization and core-periphery dynamics of (far right) platform ecosystems.

By José Van Dijk, Tim de Winkel and @mtschaefer.bsky.social
doi.org/10.1177/1461...

Steuert KI die Verwaltung, kann das mitunter gefährlich werden In den Niederlanden regeln Algorithmen soziale Dienste und lösten einen beispiellosen Skandal aus. Um solche Fälle zu vermeiden, braucht es ethische Grenzen

The Austrian @derstandard.at features @mtschaefer.bsky.social and our work on data & AI ethics (e.g. Fundamental Rights & Algorithms Impact Assessment) & how the Netherlands try to develop good governance for algorithms & AI. #Dighum2025

www.derstandard.at/story/300000...

DPG Media and Utrecht University’s Data School develop tool to support ethical use of data and AI in the media sector | Centre for Digital Humanities DPG Media and the Utrecht University Data School are jointly developing a ‘Data Ethics Decision Aid for Media (DEDA for Media)’. Starting in 2026, this analogue model will help media companies in reco...

Fascinating to see how our Data Ethics Decision Aid (DEDA) is spreading internationally & through different domains of application. We're currently working on a DEDA for media organisations, and are also developing one for robotic systems.
#AI #Ethics #Governance
cdh.uu.nl/2025/06/02/d...

Generative AI, Media, and Society | Katalin Feher | Taylor & Francis e This groundbreaking book demystifies generative AI’s transformative impact on media, socio-cultural dynamics, ethics, and policy. Defining generative AI as an

Find more information about the book here: www.taylorfrancis.com/books/mono/1...

"Generative AI, Media & Society" by Katalin Feher We are very proud to appear as one of the case studies, highlighting our work on responsible data practices & AI governance, featuring our Fundamental Rights & Algorithms Impact Assesment (FRAIA) and our Data Ethics Decision Aid (DEDA).

"Cop, Code and Conduct" by Isabelle Donatz-Fest is a deep-dive into the discourses and practices of technological change at the Dutch National Police, and a sharp analysis of an institution grappling with ethics, capacities and governance.
#AI #Policing

research-portal.uu.nl/en/publicati...

"Introducing FRAIA: The Fundamental Rights and Algorithms Impact Assessment"

Organised by @dataschool.nl with Julia Straatman, Iris Muis

More information: cpdp.be/27623
#CPDP.ai #CPDP2025 #TheWorldisWatching #CPDP #Workshop

It is this time of the year again: @cpdpconferences.bsky.social our most favorite conference to attend. We will be there with a FRAIA workshop and a panel in which we discuss how to make the AI Act work beyond the trickle-down policy, through effective practices and oversight.