The Importance of Data Access Regimes for Artificial Intelligence and Machine Learning

Curated on Posted on by Stefaan Verhulst

JRC Digital Economy Working Paper by Bertin Martens: “Digitization triggered a steep drop in the cost of information. The resulting data glut created a bottleneck because human cognitive capacity is unable to cope with large amounts of information. Artificial intelligence and machine learning (AI/ML) triggered a similar drop in the cost of machine-based decision-making and helps in overcoming this bottleneck. Substantial change in the relative price of resources puts pressure on ownership and access rights to these resources. This explains pressure on access rights to data. ML thrives on access to big and varied datasets. We discuss the implications of access regimes for the development of AI in its current form of ML. The economic characteristics of data (non-rivalry, economies of scale and scope) favour data aggregation in big datasets. Non-rivalry implies the need for exclusive rights in order to incentivise data production when it is costly. The balance between access and exclusion is at the centre of the debate on data regimes. We explore the economic implications of several modalities for access to data, ranging from exclusive monopolistic control to monopolistic competition and free access. Regulatory intervention may push the market beyond voluntary exchanges, either towards more openness or reduced access. This may generate private costs for firms and individuals. Society can choose to do so if the social benefits of this intervention outweigh the private costs.

We briefly discuss the main EU legal instruments that are relevant for data access and ownership, including the General Data Protection Regulation (GDPR) that defines the rights of data subjects with respect to their personal data and the Database Directive (DBD) that grants ownership rights to database producers. These two instruments leave a wide legal no-man’s land where data access is ruled by bilateral contracts and Technical Protection Measures that give exclusive control to de facto data holders, and by market forces that drive access, trade and pricing of data. The absence of exclusive rights might facilitate data sharing and access or it may result in a segmented data landscape where data aggregation for ML purposes is hard to achieve. It is unclear if incompletely specified ownership and access rights maximize the welfare of society and facilitate the development of AI/ML…(More)”

Data Trusts: More Data than Trust? The Perspective of the Data Subject in the Face of a Growing Problem

Curated on Posted on by Stefaan Verhulst

Paper by Christine Rinik: “In the recent report, Growing the Artificial Intelligence Industry in the UK, Hall and Pesenti suggest the use of a ‘data trust’ to facilitate data sharing. Whilst government and corporations are focusing on their need to facilitate data sharing, the perspective of many individuals is that too much data is being shared. The issue is not only about data, but about power. The individual does not often have a voice when issues relating to data sharing are tackled. Regulators can cite the ‘public interest’ when data governance is discussed, but the individual’s interests may diverge from that of the public.

This paper considers the data subject’s position with respect to data collection leading to considerations about surveillance and datafication. Proposals for data trusts will be considered applying principles of English trust law to possibly mitigate the imbalance of power between large data users and individual data subjects. Finally, the possibility of a workable remedy in the form of a class action lawsuit which could give the data subjects some collective power in the event of a data breach will be explored. Despite regulatory efforts to protect personal data, there is a lack of public trust in the current data sharing system….(More)”.

Illuminating Big Data will leave governments in the dark

Curated on Posted on by Stefaan Verhulst

Robin Wigglesworth in the Financial Times: “Imagine a world where interminable waits for backward-looking, frequently-revised economic data seem as archaically quaint as floppy disks, beepers and a civil internet. This fantasy realm may be closer than you think.

The Bureau of Economic Analysis will soon publish its preliminary estimate for US economic growth in the first three months of the year, finally catching up on its regular schedule after a government shutdown paralysed the agency. But other data are still delayed, and the final official result for US gross domestic product won’t be available until July. Along the way there are likely to be many tweaks.

Collecting timely and accurate data are a Herculean task, especially for an economy as vast and varied as the US’s. But last week’s World Bank-International Monetary Fund’s annual spring meetings offered some clues on a brighter, more digital future for economic data.

The IMF hosted a series of seminars and discussions exploring how the hot new world of Big Data could be harnessed to produce more timely economic figures — and improve economic forecasts. Jiaxiong Yao, an IMF official in its African department, explained how it could use satellites to measure the intensity of night-time lights, and derive a real-time gauge of economic health.

“If a country gets brighter over time, it is growing. If it is getting darker then it probably needs an IMF programme,” he noted. Further sessions explored how the IMF could use machine learning — a popular field of artificial intelligence — to improve its influential but often faulty economic forecasts; and real-time shipping data to map global trade flows.

Sophisticated hedge funds have been mining some of these new “alternative” data sets for some time, but statistical agencies, central banks and multinational organisations such as the IMF and the World Bank are also starting to embrace the potential.

The amount of digital data around the world is already unimaginably vast. As more of our social and economic activity migrates online, the quantity and quality is going to increase exponentially. The potential is mind-boggling. Setting aside the obvious and thorny privacy issues, it is likely to lead to a revolution in the world of economic statistics. …

Yet the biggest issues are not the weaknesses of these new data sets — all statistics have inherent flaws — but their nature and location.

Firstly, it depends on the lax regulatory and personal attitudes towards personal data continuing, and there are signs of a (healthy) backlash brewing.

Secondly, almost all of this alternative data is being generated and stored in the private sector, not by government bodies such as the Bureau of Economic Analysis, Eurostat or the UK’s Office for National Statistics.

Public bodies are generally too poorly funded to buy or clean all this data themselves, meaning hedge funds will benefit from better economic data than the broader public. We might, in fact, need legislation mandating that statistical agencies receive free access to any aggregated private sector data sets that might be useful to their work.

That would ensure that our economic officials and policymakers don’t fly blind in an increasingly illuminated world….(More)”.

Data Collaboratives as an enabling infrastructure for AI for Good

Curated on Posted on by Stefaan Verhulst

Blog Post by Stefaan G. Verhulst: “…The value of data collaboratives stems from the fact that the supply of and demand for data are generally widely dispersed — spread across government, the private sector, and civil society — and often poorly matched. This failure (a form of “market failure”) results in tremendous inefficiencies and lost potential. Much data that is released is never used. And much data that is actually needed is never made accessible to those who could productively put it to use.

Data collaboratives, when designed responsibly, are the key to addressing this shortcoming. They draw together otherwise siloed data and a dispersed range of expertise, helping match supply and demand, and ensuring that the correct institutions and individuals are using and analyzing data in ways that maximize the possibility of new, innovative social solutions.

Roadmap for Data Collaboratives

Despite their clear potential, the evidence base for data collaboratives is thin. There’s an absence of a systemic, structured framework that can be replicated across projects and geographies, and there’s a lack of clear understanding about what works, what doesn’t, and how best to maximize the potential of data collaboratives.

At the GovLab, we’ve been working to address these shortcomings. For emerging economies considering the use of data collaboratives, whether in pursuit of Artificial Intelligence or other solutions, we present six steps that can be considered in order to create data collaborative that are more systematic, sustainable, and responsible.

The need for making Data Collaboratives Systematic, Sustainable and Responsible
  • Increase Evidence and Awareness…
  • Increase Readiness and Capacity…
  • Address Data Supply and Demand Inefficiencies and Uncertainties…
  • Establish a New “Data Stewards” Function…
  • Develop and strengthen policies and governance practices for data collaboration…

Safeguards for human studies can’t cope with big data

Curated on Posted on by Stefaan Verhulst

Nathaniel Raymond at Nature: “One of the primary documents aiming to protect human research participants was published in the US Federal Register 40 years ago this week. The Belmont Report was commissioned by Congress in the wake of the notorious Tuskegee syphilis study, in which researchers withheld treatment from African American men for years and observed how the disease caused blindness, heart disease, dementia and, in some cases, death.

The Belmont Report lays out core principles now generally required for human research to be considered ethical. Although technically governing only US federally supported research, its influence reverberates across academia and industry globally. Before academics with US government funding can begin research involving humans, their institutional review boards (IRBs) must determine that the studies comply with regulation largely derived from a document that was written more than a decade before the World Wide Web and nearly a quarter of a century before Facebook.

It is past time for a Belmont 2.0. We should not be asking those tasked with protecting human participants to single-handedly identify and contend with the implications of the digital revolution. Technological progress, including machine learning, data analytics and artificial intelligence, has altered the potential risks of research in ways that the authors of the first Belmont report could not have predicted. For example, Muslim cab drivers can be identified from patterns indicating that they stop to pray; the Ugandan government can try to identify gay men from their social-media habits; and researchers can monitor and influence individuals’ behaviour online without enrolling them in a study.

Consider the 2014 Facebook ‘emotional contagion study’, which manipulated users’ exposure to emotional content to evaluate effects on mood. That project, a collaboration with academic researchers, led the US Department of Health and Human Services to launch a long rule-making process that tweaked some regulations governing IRBs.

A broader fix is needed. Right now, data science overlooks risks to human participants by default….(More)”.

Data Cultures, Culture as Data

Curated on Posted on by Stefaan Verhulst

Introduction to Special Issue of Cultural Analytics by Amelia Acker and Tanya Clement: “Data have become pervasive in research in the humanities and the social sciences. New areas, objects, and situations for study have developed; and new methods for working with data are shepherded by new epistemologies and (potential) paradigm shifts. But data didn’t just happen to us. We have happened to data. In every field, scholars are drawing boundaries between data and humans as if making meaning with data is innocent work. But these boundaries are never innocent. Questions are emerging about the relationships of culture to data—urgent questions that focus on the codification (or code-ification) of social and cultural bias and the erosion of human agency, subjectivity, and identity.

For this special issue of Cultural Analytics we invited submissions to respond to these concerns as they relate to the proximity and distance between the creation of data and its collection; the nature of data as object or content; modes and contexts of data circulation, dissemination and preservation; histories and imaginary data futures; data expertise; data and technological progressivism; the cultivation and standardization of data; and the cultures, communities, and consciousness of data production. The contributions we received ranged in type from research or theory articles to data reviews and opinion pieces responding to the theme of “data cultures”. Each contribution asks questions we should all be asking: What is the role we play in the data cultures/culture as data we form around sociomaterial practices? How can we better understand how these practices effect, and affect, the materialization of subjects, objects, and the relations between them? How can we engage our data culture(s) in practical, critical, and generative ways? As Karen Barad writes, “We are responsible for the world in which we live not because it is an arbitrary construction of our choosing, but because it is sedimented out of particular practices that we have a role in shaping.”1Ultimately, our contributors are focused on this central concern: where is our agency in the responsibility of shaping data cultures? What role can scholarship play in better understanding our culture as data?…(More)”.

Digital Health Data And Information Sharing: A New Frontier For Health Care Competition?

Curated on Posted on by Stefaan Verhulst

Paper by Lucia Savage, Martin Gaynor and Julie Adler-Milstein: “There are obvious benefits to having patients’ health information flow across health providers. Providers will have more complete information about patients’ health and treatment histories, allowing them to make better treatment recommendations, and avoid unnecessary and duplicative testing or treatment. This should result in better and more efficient treatment, and better health outcomes. Moreover, the federal government has provided substantial incentives for the exchange of health information. Since 2009, the federal government has spent more than $40 billion to ensure that most physicians and hospitals use electronic health records, and to incentivize the use of electronic health information and health information exchange (the enabling statute is the Health Information Technology for Clinical Health Act), and in 2016 authorized substantial fines for failing to share appropriate information.

Yet, in spite of these incentives and the clear benefits to patients, the exchange of health information remains limited. There is evidence that this limited exchange in due in part to providers and platforms attempting to retain, rather than share, information (“information blocking”). In this article we examine legal and business reasons why health information may not be flowing. In particular, we discuss incentives providers and platforms can have for information blocking as a means to maintain or enhance their market position and thwart competition. Finally, we recommend steps to better understand whether the absence of information exchange, is due to information blocking that harms competition and consumers….(More)”

Synthetic data: innovation for public good

Curated on Posted on by Stefaan Verhulst

Blog Post by Catrin Cheung: “What is synthetic data, and how can it be used for public good? ….Synthetic data are artificially generated data that have the look and structure of real data, but do not contain any information on individuals. They also contain more general characteristics that are used to find patterns in the data.

They are modelled on real data, but designed in a way which safeguards the legal, ethical and confidentiality requirements of the original data. Given their resemblance to the original data, synthetic data are useful in a range of situations, for example when data is sensitive or missing. They are used widely as teaching materials, to test code or mathematical models, or as training data for machine learning models….

There’s currently a wealth of research emerging from the health sector, as the nature of data published is often sensitive. Public Health England have synthesised cancer data which can be freely accessed online. NHS Scotland are making advances in cutting-edge machine learning methods such as Variational Auto Encoders and Generative Adversarial Networks (GANs).

There is growing interest in this area of research, and its influence extends beyond the statistical community. While the Data Science Campus have also used GANs to generate synthetic data in their latest research, its power is not limited to data generation. It can be trained to construct features almost identical to our own across imagery, music, speech and text. In fact, GANs have been used to create a painting of Edmond de Belamy, which sold for $432,500 in 2018!

Within the ONS, a pilot to create synthetic versions of securely held Labour Force Survey data has been carried out using a package in R called “synthpop”. This synthetic dataset can be shared with approved researchers to de-bug codes, prior to analysis of data held in the Secure Research Service….

Although much progress is done in this field, one challenge that persists is guaranteeing the accuracy of synthetic data. We must ensure that the statistical properties of synthetic data match properties of the original data.

Additional features, such as the presence of non-numerical data, add to this difficult task. For example, if something is listed as “animal” and can take the possible values “dog”,”cat” or “elephant”, it is difficult to convert this information into a format suitable for precise calculations. Furthermore, given that datasets have different characteristics, there is no straightforward solution that can be applied to all types of data….particular focus was also placed on the use of synthetic data in the field of privacy, following from the challenges and opportunities identified by the National Statistician’s Quality Review of privacy and data confidentiality methods published in December 2018….(More)”.

Credit denial in the age of AI

Curated on Posted on by Stefaan Verhulst

Paper by Aaron Klein: “Banks have been in the business of deciding who is eligible for credit for centuries. But in the age of artificial intelligence (AI), machine learning (ML), and big data, digital technologies have the potential to transform credit allocation in positive as well as negative directions. Given the mix of possible societal ramifications, policymakers must consider what practices are and are not permissible and what legal and regulatory structures are necessary to protect consumers against unfair or discriminatory lending practices.

In this paper, I review the history of credit and the risks of discriminatory practices. I discuss how AI alters the dynamics of credit denials and what policymakers and banking officials can do to safeguard consumer lending. AI has the potential to alter credit practices in transformative ways and it is important to ensure that this happens in a safe and prudent manner….(More)”.

Digital Data for Development

Curated on Posted on by Stefaan Verhulst

LinkedIn: “The World Bank Group and LinkedIn share a commitment to helping workers around the world access opportunities that make good use of their talents and skills. The two organizations have come together to identify new ways that data from LinkedIn can help inform policymakers who seek to boost employment and grow their economies.

This site offers data and automated visuals of industries where LinkedIn data is comprehensive enough to provide an emerging picture. The data complements a wealth of official sources and can offer a more real-time view in some areas particularly for new, rapidly changing digital and technology industries.

The data shared in the first phase of this collaboration focuses on 100+ countries with at least 100,000 LinkedIn members each, distributed across 148 industries and 50,000 skills categories. In the near term, it will help World Bank Group teams and government partners pinpoint ways that developing countries could stimulate growth and expand opportunity, especially as disruptive technologies reshape the economic landscape. As LinkedIn’s membership and digital platforms continue to grow in developing countries, this collaboration will assess the possibility to expand the sectors and countries covered in the next annual update.

This site offers downloadable data, visualizations, and an expanding body of insights and joint research from the World Bank Group and LinkedIn. The data is being made accessible as a public good, though it will be most useful for policy analysts, economists, and researchers….(More)”.