Data Stewardship Trends to Watch

Recapping Monthly Developments in a Changing Data Ecosystem, Non Traditional Data, AI Readiness and 4th Wave of Open Data, Value vs Extraction and more

Led by Stefaan Verhulst, The Data Tank and The GovLab organized the fall session of the Data Stewardship Trends to Watch series.

Trends to Watch is a bi-monthly series of webinars organized exclusively for our growing global data stewards alumni to stay updated on the emerging trends in the fields of data stewardship and AI.

The September session focused on non-traditional data (smart data), AI readiness and fourth wave of open data, the importance of value amidst AI induced data extraction approaches and the emergence of new tools that can be used to further the common good. In the end, Mark Herringer, a data steward alumni leading the healthsites.io  (a recognized digital public good) initiative briefly presented his important work in the health space.

Check out a recording of the full conversation here. The following resources were discussed during the session:

Trends:

Data Stewardship and Non Traditional Data

• Smart Data Challenge is an initiative undertaken by the UK Government on investing in Smart Data. The Challenge asked a simple question: if people and businesses could share their data securely, what better services could appear? Finalists developed solutions to help households manage energy and reduce bills, support small businesses with costs and reporting, simplify homebuying, and make complex markets easier to navigate.

• Do Online Job Postings Capture Job Vacancies? is a working paper published by the National Bureau of Economic Research that tries to understand the biases in non-traditional data (i.e. representativeness). Researchers question the extent to which you can use job openings as a proxy for determining job markets. The report reveals a type of bias in that data as the results are skewed towards big corporations who tend to post more job openings.

• Enabling independent research on how people use Claude is a pilot that Anthropic ran to understand Claude prompts and use them as proxies for human behaviors. This is relevant in light of non-traditional data sources as prompts could classify as non-traditional data sources similar to sentiment analysis (or search information) was used before.

4th Wave of Open Data and AI Readiness

• U.S. National Science foundation launches the Open Knowledge Network is a recently new national open-data platform that connects more than 40 knowledge graphs to power the new wave of AI-enabled discovery. Across science, government, and many key enterprise applications, valuable datasets sit in silos that are not accessible, interoperable, nor ready for use by AI systems. OKN makes federal and public data findable, connected, and machine-readable at a national scale.

• The World Bank published the World Development Report 2026 that provides the first comprehensive assessment of what AI means for economies - how firms and governments are already using it, what is holding them back, where the largest gains may lie, and how the risks can be managed. What’s striking is that data quality and availability remain one of the most common barriers to AI use in government which links to a reemphasized need for data stewardship.

• Data Harnessed for AI-Ready Advancements is an initiative that enables institutions to harmonise data, build trust for sharing across silos, optimise their data workflows with AI-ready tools and playbooks and makes these processes repeatable at scale.

Value vs Extraction

• The People’s Republic of Data: How China Is Turning Data into Capital is an issue that addresses China’s data policies that enabled turning data into capital. It outlines China’s data strategy and industrial policy, commercialization drive that enables strict state oversight of data and some of China’s successes (such as the adoption of official data exchange infrastructure and published use cases for data applications).

• Indigenous Data Sovereignty and the Governance of Artificial Intelligence: Toward Equitable Benefit for Indigenous Peoples is an article that argues that artificial intelligence (AI) will not benefit Indigenous peoples equitably unless grounded in Indigenous data sovereignty and Indigenous governance of AI systems. In Australia and comparable nation-states, data ecosystems have long served as instruments of colonial control, determining what knowledge is valued, how people are categorized, and who benefits. As AI systems are built upon these foundations, they risk deepening and automating colonial logics. Indigenous data sovereignty offers a powerful counter framework, asserting the rights of Indigenous peoples to govern the creation, interpretation, and use of data concerning themselves, their communities, and their lands. This article examines how applying Indigenous governance to data and AI systems could alleviate unintended harms and create more equitable benefits within digital technologies. Rather than treating AI as neutral or benevolent, it is positioned as a contested domain where Indigenous resurgence confronts colonial continuity, advancing governance frameworks enabling just, accountable, and equitable futures.

• Navigating Indigenous Data Sovereignty in the Age of AI is a blog that bulbs on the previous News Commons Incubator Webinar that addresses the lines mentioned in the previous article.

• AI models need more data about biology, and OpenAI is paying to create it is a piece that reports on a new effort from OpenAI Foundation that is called Public Data for Health, which aims to help artifical intelligence make big leaps in medicine by paying to create “high-quality scientific datasets”.

• Should Governments charge for public data? is an article that reports on the UK's discussions and reconsideration of funding public data. This falls on a potential 4th wave of open data as we see that the open data space is evolving. The author proposes four aspects of the UK's proposals that are critical to get right.

• Value in the emerging AI-mediated information ecosystem is a report that synthesizes four Signals at Scale summits involving approximately 140 participants from technology, investment, governance, academia, media, and adjacent fields. Its starting premise is that AI will fundamentally reorganize how societal information is observed, produced, represented, verified, distributed, experienced, acted upon, and rewarded. The emergence of AI requires nothing less than a fundamental redesign of society's knowledge infrastructure and of the information flows through which people and institutions perceive reality, form judgments, coordinate, and exercise power.

Tools

As part of the discussion, a set of tools were shared that can be useful to data and AI professionals. See below:

• Health Data Stewardship Implementation Toolkit published by the Canadian Institute for Health Information;

• The Theory of Change Process - Guidance for Outcome Delivery Plans published by the Government Analysis Function

• Academic Engagement Toolkit published by the UK Government

• Data Governance Self-Assessment Toolkit published by the GovLab and Unesco.

Recent Writings and Updates

To wrap up the trends to watch, we usually summarize recent internal updates and writings. See below:

• From Voice to Leverage in Data Governance is a paper that argues that the data governance field needs to complement voice with leverage - the capacity of individuals and communities to make their participation consequential by attaching credible legal, institutional, contractual, economic, or reputational costs to being ignored.

• Who decides what and how data is reused? Lessons learned from youth engagement and co-design for responsible data reuse in services is an article that presents findings from the NextGenData project, which developed and piloted a scalable methodology for engaging young people aged 19–24 in co-designing responsible data reuse strategies. Conducted as a year-long participatory action research initiative across India, Tanzania, Moldova, and Kyrgyzstan, the approach implemented youth assemblies, deliberative methods, and localised facilitation by national partners to engage young people.

• A roadmap for digital self-determination in data spaces is a piece that seeks to provide an implementable architecture for operationalizing Digital Self-Determination within Switzerland’s data ecosystem, with a particular focus on trustworthy data spaces as the primary site of action. It is intended for data space architects, governance authorities, and organizations or individuals responsible for designing and operating data sharing infrastructures. It builds directly on existing Swiss policy instruments, including the Federal Council report Creating trustworthy data spaces based on digital self-determination and the Code of Conduct for operating trustworthy data spaces based on digital self-determination, and focuses on how their principles can be implemented in practice.

• Rethinking Official Statistics Beyond Averages: Centering Lived Experience is a conference paper that tackles the landscape official statistics are operating in. This paper argues that the crisis is not primarily one of methodological failure or statistical illiteracy, but of representational legitimacy: aggregate indicators systematically fail to capture lived experience, and citizens increasingly do not recognize themselves in the numbers that purport to describe them. We situate this argument within a broader body of work on moving “from averages to agency”.

Interested to learn more about data stewardship trends? Sign up for the Data Stewards Network mailing at this link. Or contact contact@datastewards.net to learn more.