iiinova — where your digital self keeps evolving

And with that, my train pulls back into Delhi. I'll be joining the opening sessions of @pairs.site online from.our hub here on Tuesday. If any of the talks pique your interest (they will be put more accessibly than my short/verbose skeets) do join us via Zoom: details at pairs.site

An image of a station through a grubby train window as the dusk falls. Not quite the approach to Delhi: it's already too dark. I promised a thread of high quality participatory AI insight and low quality transit photos. I hope I have delivered. Please complete the feedback form on your seat and leave it on the desk as you leave.

000

If the papers and slides already shared are anything to go by, PAIRS 2026 Online will be an intense day. But as intense as a week of TukTuk rides and Delhi traffic. I recon it might be a close run thing for anyone who does a full day of PAIRS online: but you can just dip in and out too!

Image of TukTuk in traffic. I didn't capture photos of the really intense bits as I was holding on tight rather than taking shots with my phone.

100

Just waiting to board return from @pairs.site program committee outing to Agra, so online conference paper previews (and poor quality travel photography) to resume shortly.

121

And as we're nearly in Agra I'll pause this thread for now, offering you as promised another compelling snapshot from a moving train.

100

Also in the parallel session at 9.45 UTC (Zoom Rm B) Chiara Ullstein will interrogate how far the EU AI Act provides a mandate for participatory practice (Spoiler: not as much as you would hope, but there are some regulatory hooks in there...). www.pairs.site/Participator...

100

Tim Davies @timdavies.org.uk · 14/02/2026

Back with some more 10 minute talks not to miss at the free PAIRS Online conference on Zoom on Tuesday as I get the express train down to Agra for a visit to the Taj Mahal in what I can now promise will be a thread of People power & AI, combined with poor quality photos from moving vehicles...

Picture of e-reader and paper on participatory AI design justice, taken looking out (rather dusty) window of an Indian express train. Tracks, trash and tower blocks complete the view.

311

Ok. Should have realised I'm Delhi traffic/taxi ride not super conducive to transcribing paper notes from the plane. Will share more later. For now, a poor quality photo of one of the #AIImpactSummit hoardings outside the airport

AI impact Summit banner. accelerating AI for social and global good. 3 guiding sutras. People, planet, progress

100

Tim Davies @timdavies.org.uk · 12/02/2026

On Tuesday (still some spaces left) we'll have a joint roundtable with @pairs.site and the Cyril Shroff Centre to look at how Right to Information, Data Protection and other legal foundations of democratic governance fit into the picture of participatory AI. www.pairs.site/New-Delhi-Si...

Logos of PAIRS and Jindal Global Law School

100

Tim Davies @timdavies.org.uk · 30/01/2026

Voice isn't an add-on element of the framework, but at it's core: "...an essential prerequisite for ethical and effective information use that the multiple voices of children, young people, families and practitioners are heard both within and about their information and its use." (3/4)

It is an essential prerequisite for ethical and
effective information use that the multiple
voices of children, young people, families
and practitioners are heard both within
and about their information and its use.
Their voices need to be integrated within
children’s information at the aggregate
and individual levels, in ways that are
meaningful and that can be accessed,
analysed, understood and used.

100

Tim Davies @timdavies.org.uk · 30/01/2026

Encouraging to see practices focussed on the voice of children, young people and families central to latest report from @nuffieldfoundation.org Children's Information Project. www.childrensinformationproject.org.uk/report (1/4)

A circular infographic titled "Figure 1: A Framework for the implementation of ethical and effective information use."

The diagram is organized into three concentric layers:

Outer Ring: Divided into two semi-circles labeled Infrastructure and Governance (top-right) and Behaviours and Culture (bottom-left).

Middle Ring: A color-coded wheel with eight segments detailing specific actions:

Infrastructure & Governance side: "Bring voice into co-design of information systems" (orange), "Using theory of change" (light orange), "Mapping information" (red), and "Mapping systems" (purple).

Behaviours & Culture side: "Drawing on national datasets" (lime green), "Improving voice in operational information use" (mint green), "Integrating children's and families' voices into information use" (teal), and "Using broad sources of local information" (blue).

Inner Core: A smaller circle divided by a cross into four quadrants, labeled clockwise as Best use of information, Integrating Voice, Understanding Needs, and Appropriate Action.

Arrows indicate a continuous, cyclical flow between the segments and rings.

130

Tim Davies @timdavies.org.uk · 28/01/2026

The hard law foundations of participatory AI is a theme I'm expecting to come up in quite a few papers @pairs.site in Feb - and reminding myself of the work already done to set out this in Michelle's paper: datasociety.net/library/demo...

8: Mandate public participation with “hard law” requirements
with concrete enforcement mechanisms.

It is not enough for public participation to be advisory or recommended—or delivered by the benevolence of corporate entities. Rather, hard law requirements mandating public participation and creating enforceable rights to participation are foundational to ensuring that decision-makers take participatory requirements seriously and that the public has real
influence.

Hard law requirements ensure continuity within organizations over time, help build institutional norms around participation, and increase trust with affected communities. An ideal statute for public participation in AI would incorporate the preceding guidelines by:

• mandating public participation;
• defining meaningful participation and the goals of participation;
• describing the public’s role at various stages of the AI life cycle;
• requiring two-way forms of deliberation at specific stages;
• directing culturally appropriate outreach to affected communities;
• identifying obligations to increase access for marginalized communities;
• committing funds for public participation and specialized support on technological issues; and
• setting forth enforcement mechanisms to ensure compliance with public participation requirements.

022

Tim Davies @timdavies.org.uk · 27/01/2026

With growing calls for public deliberation on AI (e.g. this exciting grassroots campaign launching next week: us02web.zoom.us/meeting/regi... ) this is key insight: big debates about AI governance will be more powerful if linked to ground-level experience, and proxies through which AI is encountered

100

Tim Davies @timdavies.org.uk · 27/01/2026

This resonates with our work on public deliberation on AI: when we move from abstract discussions, to concrete experiences of AI-backed systems, we can both help public dialogue grasp hold of AI, and highlight opportunities for agency.

Photos from the Hopes and Fears Lab that took place as part of the 2023 People's Panel on AI.

Hopes and Fears lab format by Kavli Centre.

100

Tim Davies @timdavies.org.uk · 27/01/2026

Through questions and methods that sensitise publics and researchers to the presence of data and AI systems in the mixed public-private space of the contemporary street, the project highlights that citizens mostly encounter AI through proxies.

Figure from AI in the Streets report showing 'Types of AI and AI 'proxies' identified in the five study locations (by category).'

Including images of phones, drones, bikes, traffic lights, screens and sensors.

Figure from AI in the Streets report showing 'Types of AI and AI 'proxies' identified in the five study locations (by category).'

Labelled bubbles for 'AI-enabling infrastructure', 'AI agentis', 'AI infrastructure', 'AI Applications' and 'Other'

100

Tim Davies @timdavies.org.uk · 23/01/2026

Just written to @simonforstroud.bsky.social to ask support for Early Day Motion 2650 calling for a 'UK digital sovereignty strategy' based on #openStandards & #interoperability: We need "ability to maintain, modify, and replace [our tech] without dependence on a single supplier

UK digital sovereignty strategy
EDM (Early Day Motion)2650: tabled on 20 January 2026
Tabled in the 2024-26 session.

This motion has been signed by 3 Members. It has not yet had any amendments submitted.

Motion text
That this House notes that government services, democratic functions and critical infrastructure increasingly depend on a small number of external digital suppliers; further notes that excessive concentration and inadequate exit or substitution planning expose the public sector to risks including service withdrawal, sanctions, commercial failure, geopolitical disruption and unilateral changes in service terms; also notes that existing government policy, including the Technology Code of Practice and Open Standards Principles, promotes interoperability, portability, reuse, access to source code and open source approaches to reduce supplier dependency, but that these principles are not applied consistently or in a binding way across government; believes that long-term resilience, continuity of public services and value for money require the Government to retain effective control over digital systems it funds or relies upon, including the ability to maintain, modify, and replace them without dependence on a single supplier; further believes that reducing supplier concentration through interoperable systems would strengthen competition, support UK technology firms and SMEs, and increase the proportion of public digital expenditure retained in the UK economy; and calls on the Government to publish a comprehensive UK digital sovereignty strategy with binding effect across central government, arm’s-length bodies, and the wider public sector, setting out how continuity of service and operational autonomy will be ensured in the event of supplier withdrawal, sanctions or external state action.

111

Join @connectedbydata.org's Webinar on April 22nd when we'll talk more about learning from Generative AI in Education: Have Your Say, and thinking about ways for students to have a powerful voice in decisions about data & AI from classroom to capital. connectedbydata.org/events/2026-... (14/15)

121

We heard across the distributed dialogue that students want to see tools proven as safe, accurate and fair before they are deployed. They expect those assurances from the government, but they also want students to be listened to and involved in those decisions. (12/15)

Take the time to get generative AI in education right.

Students want assurance that AI is accurate, and that bias and safety risks of AI are well addressed before it is deployed in the classroom. Focus on limited classroom and homework uses of AI, rather than AI everywhere.

The Department for Education’s updated “Generative AI: product safety standards” reflect many of the concerns raised by students during the distributed dialogue.

More at https://www.gov.uk/government/publications/generative-ai-product-safety-standards

Embed a role for students in decision-making on AI.

AI for education should be co-designed with student input. Support should be available for teachers and education leaders to have ongoing and informed dialogue with students about shared expectations around AI in education.

The ‘workshop in a box’ remains available to support local school dialogues. Connected by Data can also provide advice to companies on involving affected communities in tool design and governance.

More at https://connectedbydata.org/ai-in-education/

Build student agency in a world of AI.

Young people hold a diversity of attitudes towards AI and its future use. Support students to learn about the pros and cons of particular uses of AI, and to make informed individual and collective choices about when and how they engage with it.

Many teachers told us they were surprised by the views students expressed during workshops. Students, teachers and parents all need regular opportunities to discuss the use of Generative AI in Education.

More at https://connectedbydata.org/ai-in-education/

However, it is great to see such a strong connection between the views students shared in our distributed dialogues connectedbydata.org/ai-in-educat... and the updated product safety expectations. The real test though now is whether these expectations can be operationalised. (10/15)

Results image showing a graph of student 'top and bottom tools' for Gen AI in Education.

We asked students to think about different AI tools for education.

Top tools were those that put students in control of their own research, allow students to work at their own pace, and focus on personalised revision or reminders of the key things you need to learn.

“[these] would support us without feeling like it was cheating or we weren't learning the information.“

AI for lesson planning and marking featured most commonly in the ‘bottom three’ of the rankings.

“Work is something we put a lot of effort into and we feel it needs to be read by a teacher.“

Image showing student drawings about AI and text: "Creative feedback demonstrated how students feel about the importance of the student-teacher relationship, using AI to make lessons more engaging, and to support personalised and self-directed learning.
Students want to be prepared to navigate a future world where AI may be ubiquitous, but where they can make choices about how to use it. "

Groups voted on a set of statements about Generative AI in Education.
Students could also suggest their own statements for others to vote on.
We used the pol.is platform that creates distinct clusters of respondents and statements based on (a) bringing together groups that responded in similar ways to statements; (b) separating out groups that responded differently.

Cluster 1: AI sceptics
24 groups – 49%

This cluster said: AI use should not be compulsory; there should be restrictions on data sharing with AI, and on where AI is used.

Cluster 2: Trusting users
16 groups – 33%

This cluster said: AI must be accurate, and AI tools tailored for educational use. Then it can be a useful part of learning.

Cluster 3: Self-assured users
9 groups – 18%

This cluster said students should be allowed or encouraged to use AI and should take responsibility for being aware of its limitations.

121

The cognitive development expectations capture idea of a ‘schools mode’ that pupils raised in GenAI in Education dialogues. DfE say “We expect products not to provide final answers, full solutions, or complete worked examples by default”, but instead to help students learn step-by-step. (4/15)

Screenshot of gov.uk page.

Guidance: Generative AI: Product Safety Standards - Updated 19 January 2026. Applies to England.

The last, and smallest, cluster we called the 'Self-assured users'. Students in this group felt it was ok to AI to give incorrect answers, or to sound like it has feelings, suggesting these groups felt confident that they could navigate the limits of AI when using it in education.

Slide image:

This is a small cluster, with few unanimous defining statements. However:

* it is the only cluster where a majority agree it is ok to AI to give incorrect answers, and for AI to sounds like it has feelings

* it is the only cluster to dismiss environmental concerns

* it is the cluster most likely to be positive about AI image generation
* it is the cluster most likely to object to requirements that students should tell teachers when they’ve used generative AI, and to be comfortable with AI ‘teaching to the test’

This cluster included a higher representation of independent schools. It may represent a group of students more assured of their own ability, or the ability of their teachers to support them to navigate the use of current AI systems well.

The second cluster we labelled 'Trusting users'. These students accepted that AI might become an essential tool in the classroom, but either assume, or demand, that is must be accurate. This cluster wanted to see AI tools for independent learning, and that support students to use AI responsibly.

Image of Cluster 2 slide.

This cluster voted unanimously that:

* sometimes AI use could be mandatory in class

* it is not ok for AI to give pupils incorrect information

There was strong support in this cluster to schools to have AI systems that students can use independently to support their learning, and for the idea that AI tools should support students in being accountable for their AI use.

This group either assume that AI tools are already accurate and safe, or are focussed on targeted use of AI to support particular aspects of learning.

100

The largest cluster were the 'AI Sceptics': opposed to mandatory introduction of AI in the classroom, concerned about accuracy, and supportive of limits on student use of AI. In creative feedback students shared concerns about AI driving too much screen time, or undermining satisfaction in learning

Slide from the report

Students are concerned about the mental health impacts of AI. They do not want AI tools to be involved in personal, emotional and social support. A clear majority of groups agreed that:

“AI tutors should never be able to have conversations with pupils about personal issues, like friendships or family problems.”

They opposed AI tools pretending to have feelings, though some suggested it could help if AI tools offered light encouragement.

Students were generally in favour of the statement that:

“Children from a young age should be educated on how to use AI most effectively and safely in their learning”

The slide also contains a student cartoon, and worksheet responses, sharing views on education with and without AI.

After exploring balanced learning materials, groups voted on a number of statements using the pol.is platform. While almost all groups agreed AI should be safe, accurate and fair before deployment, and should never replace face-to-face time with teachers, we found three distinct clusters of views...

Voting and Clusters Slide:

After learning about potential tools for AI in education, and exploring different issues with AI we asked student
groups to vote on a set of statements, and to suggest their own.

Statements were shown in a random order. Groups could agree, disagree, or pass if unsure.

We used the pol.is platform to record responses, and cluster results: building a picture of different student perspectives on AI in education.

Almost everyone agreed that AI tools should be officially checked for safety, accuracy and bias before deployment in education and that AI should never replace face-to-face time with teachers.

Cluster 1: AI sceptics
24 groups – 49%
AI use should not be compulsory. There should be
restrictions on data sharing with
AI, and on where AI is used.

Cluster 2: Trusting users
16 groups – 33%
AI must be accurate, and AI tools tailored for educational use. Then it can be a useful part of learning.

Cluster 3: Self-assured users
9 groups – 18%
Students should be allowed or
encouraged to use AI and should take responsibility for being aware of its limitations.

Tim Davies @timdavies.org.uk · 27/10/2025

If you've been thinking of sending in an abstract - please do! 2026 Timeline: short abstracts due Friday 31st October; Notifications by 5th December; final papers, posters or presentations due 30th Jan. www.pairs.site

**Who is PAIRS for?**

We seek to create a space that is equally open to contributions from civil society, academia and industry, and that welcomes both presentation of practice, and empirical or theoretical work.

This call is open to all applicants regardless of geographic location or institutional affiliation. However we are especially interested in receiving applications from civil society organisations based in India; South Asia; Most Affected People and Areas (MAPA); members of the Global Majority or Global South; Black, Indigenous, and other People of Color; women, transgender, non-binary, and/or gender-diverse applicants; migrant and diasporic communities; and/or persons coming from climate displaced/impacted communities, etc. We strongly encourage all such contributors to submit an abstract.

We welcome submissions that represent cross-sector and cross-country collaboration.

We will prioritise work that has not been previously published or presented in research venues.

Track 1: Participatory AI Development
How can those designing, developing and integrating AI tools and platforms involve affected publics in systems design and development?
Track 2: Participatory AI Governance
How can affected communities be meaningfully involved in global, national and local governance of AI?
Track 3: AI Participation, Power and Negotiation
How can community refusal, contestations and questioning reshape AI development or governance?

010

Tim Davies @timdavies.org.uk · 27/10/2025

⏳Call for Abstracts for PAIRS26 - the 'India AI Impact Summit edition', closes in 5 days on Fri 31st Oct. www.pairs.site/PAIRS26-Call... If you've research or practice to share on participatory approaches to AI, or you've been thinking about submitting but not got around yet, please take a look.

112

Tim Davies @timdavies.org.uk · 11/06/2025

"...we should not task machines with decisions simply because they now seem capable of making them; we should instead do so based on whether ceding those decisions expands or contracts our agency and freedoms." from HDR2025 - Chapter 2

030

Tim Davies @timdavies.org.uk · 16/04/2025

Cycled up to Hawkwood this morning to prepare Perspectives on AI course I'm running next month: a one day intro to critically engaging with AI. www.hawkwoodcollege.co.uk/our-programm... Rainy now, but with gd weather in May we'll make most of the setting to slow down & reflect on fast moving tech

View from a Stone Window into a lush green garden with trees and hillside beyond

020

Tim Davies @timdavies.org.uk · 21/03/2025

Bookending a week of dipping back into the world of standards, a copy of Standards by Pomerantz and Griffey has arrived. Pitch perfect opening: "There is a standard for standards... because of course there is." mitpress.mit.edu/978026254949...

Book cover image. Standard by Jeffrey Pomerantz and Jason Griffey.

011

Tim Davies @timdavies.org.uk · 17/03/2025

Being at QEII conference centre and seeing a sign to the Berners-Lee room gave me flash back to stressful hour when we had Sir Tim Berners-Lee launching the first edition of the Open Data Barometer at the OGP Summit in this building back in 2013. opendatabarometer.org/1stEdition/

Sign pointing to Berners-Lee room (and Dawkins, Hawking and Turing rooms)

220

Tim Davies @timdavies.org.uk · 08/02/2025

Packed room here for our panel on Participatory Development of AI with insights from different settings around the globe. With @connectedbydata.org and @datasociety.bsky.social

Lecture theatre full of people at SciencesPo

View from the back to lecture theatre full of participations.

0134

Tim Davies @timdavies.org.uk · 20/01/2025

Registration is open for our in-person fringe event of the AI Action Summit. On Sat 8th Feb, we'll be hearing from over 50 different speakers on aspects of participatory AI development, governance and resistance and thinking about agendas for future research & action. Info: pairs25.notion.site

Here’s the alt text for the image:

"Promotional poster for the Participatory AI Research & Practice Symposium, part of the Week of Paris AI Action Summit. The poster has a red and cream color scheme. Main text reads: 'Participatory AI Research & Practice Symposium' with the tagline 'Shaping Systems by Shifting Power.' Event details: February 8, 9 AM - 5:30 PM, SciencesPo, 27 Rue Saint-Guillaume, 75007 Paris, France. Logos of event partners at the bottom include Connected by Data, SciencesPo, Data & Society, Aapti Institute, Digital Trade & Data Governance Hub, ISWE Foundation, and Digital Good Network. The web address 'pairs25.notion.site' is at the top right."

0122

Tim Davies @timdavies.org.uk · 17/01/2025

& on 8th Feb, ahead of the AI Action Summit in Paris, we'll be gathering at SciencesPo for part 2 of the Participatory AI Research & Practice Symposium. 60+ presenters sharing work & conversation on participatory shaping of AI systems by shifting power. Info & registration pairs25.notion.site

Poster for Participatory AI Research & Practice Symposium. Shaping Systems by Shaping Power. 8th Feb. 9am - 5.30pm. SciencesPo, 27 Rue Saint-Guillaume, 750007 Paris, France. With partner logos: SciencesPo, Connected by Data, Digital Trade and Data Governance Hub, ISWE Foundation, Data & Society, Digital Good Network and Aapti Institute.

083

Tim Davies @timdavies.org.uk · 17/01/2025

The first part of the Participatory AI Research & Practice Symposium (PAIRS25) is coming up on 30th Jan from 2pm - 4pm UTC via Zoom. 10+ presentations of cutting edge and critical perspectives on democratic public inputs to AI development and governance. Info and registration: pairs25.notion.site

World cloud of the contents of papers being presented at PAIRS25 online and in-person. Large words including public, participatory, governance, research, communities, systems, data and development. Smaller words include involvement, digital, design, stakeholders, processes, healthcare, accountability and deliberation.

293

Tim Davies @timdavies.org.uk · 13/01/2025

Pilots of AI can be good: but participatory pilots where judgements on impact and scaling involve workers & affected communities. This can be simple process: it doesn't need to slow things down. But if pilots led solely by growth mission not local need we risk undermining AI opportunities long-term.

Participatory pilots 1 - 2 - 3: 1. Evaluate AI Capabilities & Develop an Experiment; 2. Identify affected stakeholders and ethical issues; 3. Pilot in partnership.

130

Tim Davies @timdavies.org.uk · 13/01/2025

A lot of the opportunities and ideas of where AI should be used in the plan are 'supply side' driven rather than exploring and understanding public demand. We talked with LGA's AI Network last yr about giving communities voice across AI development & deployment connectedbydata.org/events/2024-...

Slide showing text "Legitimacy and trust", "Assuring inputs", "Setting outcomes", "Evaluating Models"

151

Tim Davies @timdavies.org.uk · 06/01/2025

Some fantastic insights in @globalvoices.org report on AI Narratives from the Civic Media Observatory: globalvoices.org/2024/12/23/a... exploring discussions from across the globe and how how "regional conversations often miss critical connections between AI systems, data governance, and power"

Report Cover: 2024 - Artificial Intelligence Narratives: A Global Voices Report

073

Tim Davies @timdavies.org.uk · 16/12/2024

Back from a fantastic #GlosDataDay building networks for local, creative, critical and collaborative action on data. Good to test out the Community Campaigns on Data toolkit too.

Art Gallery room with chairs, tables and around 15 people in workshop small group discussions

Paper people connected by sticks: caption of 'What was the best connection you made today and why?'

040

Tim Davies @timdavies.org.uk · 26/11/2024

Would be fantastic to see strong representation across each category from communities focussed on collective, community and public-good centred data governance. Simple application process and broad eligibility criteria (see for e.g. tech community & academic community criteria below).

Technical community eligibility criteria:

Candidates must:
• Be members of a technical community (collective groups of individuals,
or organizations who collaborate on the development, implementation,
maintenance, and governance of technical standards or digital
infrastructures and their elements) actively engaged in fields relevant to
data governance, including but not limited to data security, privacy
engineering, internet architecture, open-source development, or
information technology;
• Demonstrate expertise in addressing data governance challenges and
opportunities from a technical perspective;
• Have a proven track record of contributions to technical projects,
standards development, or community initiatives related to data
governance;
• Have experience with intergovernmental or multi-stakeholder
processes.

Academic eligibility criteria. Candidates must:
• Have demonstrated expertise in fields relevant to data governance, such
as data science, data policy, digital economy, digital platforms and
ecosystems, internet governance, or information technology;
• Have a strong and internationally recognized publication record in peerreviewed journals;
• Possess a strong understanding of data governance challenges and
opportunities;
• Have experience with intergovernmental or multistakeholder processes.

021

Tim Davies @timdavies.org.uk · 14/11/2024

📣Participatory AI Research Symposium: Call for Papers Join us on fringe of Paris AI Action Summit next year to explore case for, cases of, and critical perspectives on participatory development & governance of AI. 🌍 pairs25.notion.site 📅 30 Jan (online) 8 Feb (in-person) ✍️ Abstracts by 8 Dec!

Logos of the organising partners: SciencesPo, Connected by Data, Digital Trade and Governance Hub, ISWE, Data & Society, Digital Good Network,. Public Voices in AI, Aapti Institute.

122

Tim Davies @timdavies.org.uk · 01/10/2024

Q2: What infrastructures do we need for public *voice* in AI? The paper notes need for participatory process to direct development goals & decisions, but examples of how to achieve this are limited. I'd love to see distinct category/row in the diagram below for 'voice' (4/

Diagram 1: Examples of private and public alternatives across the AI ecosystem covering: data, compute, models, hosting, modifications, applications, tools, labour, accountability.

111

Tim Davies @timdavies.org.uk · 24/09/2024

2. It's not just about data subjects Data affects people who are not included in the dataset. In connectedbydata.org/resources/in... we explore this as 'data from me', 'data about me' and 'data that affects me' (2/5)

Image from blog post including the text:

For public sector organisations that may emerge as three questions:

“Who is this data from?” - and what rights do they have over it?
“Who is this data about” - and how might we engage with them? and
“Who does this data affect?” - and do we understand those effects and impacts?