Digital Government Model


Report by USAID: “The COVID-19 pandemic demonstrated the importance of digital government processes and tools. Governments with digital systems, processes, and infrastructure in place were able to quickly scale emergency response assistance, communications, and payments. At the same time, the pandemic accelerated many risks associated with digital tools, such as mis- and disinformation, surveillance, and the exploitation of personal data.

USAID and development partners are increasingly supporting countries in the process of adopting technologies to create public value– broadly referred to as digital government–while mitigating and avoiding risks. The Digital Government Model provides a basis for establishing a shared understanding and language on the core components of digital government, including the contextual considerations and foundational elements that influence the success of digital government investments…(More)”

How the Federal Government Buys Our Cell Phone Location Data


Article by Bennett Cyphers: “…Weather apps, navigation apps, coupon apps, and “family safety” apps often request location access in order to enable key features. But once an app has location access, it typically has free rein to share that access with just about anyone.

That’s where the location data broker industry comes in. Data brokers entice app developers with cash-for-data deals, often paying per user for direct access to their device. Developers can add bits of code called “software development kits,” or SDKs, from location brokers into their apps. Once installed, a broker’s SDK is able to gather data whenever the app itself has access to it: sometimes, that means access to location data whenever the app is open. In other cases, it means “background” access to data whenever the phone is on, even if the app is closed.

One app developer received the following marketing email from data broker Safegraph:

SafeGraph can monetize between $1-$4 per user per year on exhaust data (across location, matches, segments, and other strategies) for US mobile users who have strong data records. We already partner with several GPS apps with great success, so I would definitely like to explore if a data partnership indeed makes sense.

But brokers are not limited to data from apps they partner with directly. The ad tech ecosystem provides ample opportunities for interested parties to skim from the torrents of personal information that are broadcast during advertising auctions. In a nutshell, advertising monetization companies (like Google) partner with apps to serve ads. As part of the process, they collect data about users—including location, if available—and share that data with hundreds of different companies representing digital advertisers. Each of these companies uses that data to decide what ad space to bid on, which is a nasty enough practice on its own. But since these “bidstream” data flows are largely unregulated, the companies are also free to collect the data as it rushes past and store it for later use. 

The data brokers covered in this post add another layer of misdirection to the mix. Some of them may gather data from apps or advertising exchanges directly, but others acquire data exclusively from other data brokers. For example, Babel Street reportedly purchases all of its data from Venntel. Venntel, in turn, acquires much of its data from its parent company, the marketing-oriented data broker Gravy Analytics. And Gravy Analytics has purchased access to data from the brokers Complementics, Predicio, and Mobilewalla. We have little information about where those companies get their data—but some of it may be coming from any of the dozens of other companies in the business of buying and selling location data.

If you’re looking for an answer to “which apps are sharing data?”, the answer is: “It’s almost impossible to know.” Reporting, technical analysis, and right-to-know requests through laws like GDPR have revealed relationships between a handful of apps and location data brokers. For example, we know that the apps Muslim Pro and Muslim Mingle sold data to X-Mode, and that navigation app developer Sygic sent data to Predicio (which sold it to Gravy Analytics and Venntel). However, this is just the tip of the iceberg. Each of the location brokers discussed in this post obtains data from hundreds or thousands of different sources. Venntel alone has claimed to gather data from “over 80,000” different apps. Because much of its data comes from other brokers, most of these apps likely have no direct relationship with Venntel. As a result, the developers of the apps fueling this industry likely have no idea where their users’ data ends up. Users, in turn, have little hope of understanding whether and how their data arrives in these data brokers’ hands…(More)”.

Data saves lives: reshaping health and social care with data


UK Government Policy Paper: “…Up-to-date information about our health and care is critical to ensuring we can:

  • plan and commission services that provide what each local area needs and support effective integrated care systems
  • develop new diagnostics, treatments and insights from analysing information so the public have the best possible care and can improve their overall wellbeing
  • stop asking the public to repeat their information unnecessarily by having it available at the right time
  • assess the safety and quality of care to keep the public safe, both for their individual care and to improve guidance and regulations
  • better manage public health issues such as COVID-19, health and care disparities, and sexual health
  • help the public make informed decisions about their care, including choosing clinicians, such as through patient-reported outcome measures (PROMs) that assess the quality of care delivered from a patient’s perspective

When it comes to handling personal data, the NHS has become one of the most trusted organisations in the UK by using strict legal, privacy and security controls. Partly as a consequence of this track record, the National Data Guardian’s recent report Putting Good Into Practice found that participants were supportive of health and social care data being used for public benefit. This reflects previous polls, which show most respondents would trust the NHS with data about them (57% in July 2020 and 59% in February 2020).

During the pandemic, we made further strides in harnessing the power of data:

However, we cannot take the trust of the public for granted. In the summer of 2021, we made a mistake and did not do enough to explain the improvements needed to the way we collect general practice data. The reasons for these changes are to improve data quality, and improve the understanding of the health and care system so it can plan better and provide more targeted services. We also need to do this in a more cost-effective way as the current system using ad hoc collection processes is more expensive and inefficient, and has been criticised by the National Audit Office and the House of Commons Public Accounts Committee.

Not only did we insufficiently explain, we also did not listen and engage well enough. This led to confusion and anxiety, and created a perception that we were willing to press ahead regardless. This had the unfortunate consequence of leading to an increase in the rate of individuals opting out of sharing their data. Of course, individual members of the public have the right to opt out and always will. But the more people who opt out, the greater the risk that the quality of the data is compromised….

In this data strategy, which differs from the draft we published last year, we are putting public trust and confidence front and centre of the safe use and access to health and social care data. The data we talk about is not an abstract thing: there is an individual, a person, a name behind each piece of data. That demands the highest level of confidence. It is their data that we hold in trust and, in return, promise to use safely to provide high-quality care, help improve our NHS and adult social care, develop new treatments, and, as a result, save lives…(More)”

Prediction machines, insurance, and protection: An alternative perspective on AI’s role in production


Paper by Ajay Agrawal, Joshua S. Gans, and Avi Goldfarb: “Recent advances in AI represent improvements in prediction. We examine how decisionmaking and risk management strategies change when prediction improves. The adoption of AI may cause substitution away from risk management activities used when rules are applied (rules require always taking the same action), instead allowing for decisionmaking (choosing actions based on the predicted state). We provide a formal model evaluating the impact of AI and how risk management, stakes, and interrelated tasks affect AI adoption. The broad conclusion is that AI adoption can be stymied by existing processes designed to address uncertainty. In particular, many processes are designed to enable coordinated decisionmaking among different actors in an organization. AI can make coordination even more challenging. However, when the cost of changing such processes falls, then the returns from AI adoption increase….(More)”.

A Future Built on Data: Data Strategies, Competitive Advantage and Trust


Paper by Susan Ariel Aaronson: “In the twenty-first century, data became the subject of national strategy. This paper examines these visions and strategies to better understand what policy makers hope to achieve. Data is different from other inputs: it is plentiful, easy to use and can be utilized and shared by many different people without being used up. Moreover, data can be simultaneously a commercial asset and a public good. Various types of data can be analyzed to create new products and services or to mitigate complex “wicked” problems that transcend generations and nations (a public good function). However, an economy built on data analysis also brings problems — firms and governments can manipulate or misuse personal data, and in so doing undermine human autonomy and human rights. Given the complicated nature of data and its various types (for example, personal, proprietary, public, and so on), a growing number of governments have decided to outline how they see data’s role in the economy and polity. While it is too early to evaluate the effectiveness of these strategies, policy makers increasingly recognize that if they want to build their country’s future on data, they must also focus on trust….(More)”.

AI Ethics: Global Perspectives


New Course Modules: A Cybernetics Approach to Ethical AI Designexplores the relationship between cybernetics and AI ethics, and looks at how cybernetics can be leveraged to reframe how we think about and how we undertake ethical AI design. This module, by Ellen Broad, Associate Professor and Associate Director at the Australian National University’s School of Cybernetics, is divided into three sections, beginning with an introduction to cybernetics. Following that, we explore different ways of thinking about AI ethics, before concluding by bringing the two concepts together to understand a new approach to ethical AI design.

How should organizations put AI ethics and responsible AI into practice? Is the answer AI ethics principles and AI ethics boards or should everyone developing AI systems become experts in ethics? In An Ethics Model for Innovation: The PiE (Puzzle-solving in Ethics) Model, Cansu Canca, Founder and Director of the AI Ethics Lab, presents the model developed and employed at AI Ethics Lab: The Puzzle-solving in Ethics (PiE) Model. The PiE Model is a comprehensive and structured practice framework for organizations to integrate ethics into their operations as they develop and deploy AI systems. The PiE Model aims to make ethics a robust and integral part of innovation and enhance innovation through ethical puzzle-solving.

Nuria Oliver, Co-Founder and Scientific Director of the ELLIS Alicante Unit, presentsData Science against COVID-19: The Valencian Experience”. In this module, we explore the ELLIS Alicante Foundation’s Data-Science for COVID-19 team’s work in the Valencian region of Spain. The team was founded in response to the pandemic in March 2020 to assist policymakers in making informed, evidence-based decisions. The team tackles four different work areas: modeling human mobility, building computational epidemiological models, predictive models on the prevalence of the disease, and operating one of the largest online citizen surveys related to COVID-19 in the world. This lecture explains the four work streams and shares lessons learned from their work at the intersection between data, AI, and the pandemic…(More)”.

Dynamic World


About: “The real world is as dynamic as the people and natural processes that shape it. Dynamic World is a near realtime 10m resolution global land use land cover dataset, produced using deep learning, freely available and openly licensed. It is the result of a partnership between Google and the World Resources Institute, to produce a dynamic dataset of the physical material on the surface of the Earth. Dynamic World is intended to be used as a data product for users to add custom rules with which to assign final class values, producing derivative land cover maps.

Key innovations of Dynamic World

  1. Near realtime data. Over 5000 Dynamic World image are produced every day, whereas traditional approaches to building land cover data can take months or years to produce. As a result of leveraging a novel deep learning approach, based on Sentinel-2 Top of Atmosphere, Dynamic World offers global land cover updating every 2-5 days depending on location.
  2. Per-pixel probabilities across 9 land cover classes. A major benefit of an AI-powered approach is the model looks at an incoming Sentinel-2 satellite image and, for every pixel in the image, estimates the degree of tree cover, how built up a particular area is, or snow coverage if there’s been a recent snowstorm, for example.
  3. Ten meter resolution. As a result of the European Commission’s Copernicus Programme making European Space Agency Sentinel data freely and openly available, products like Dynamic World are able to offer 10m resolution land cover data. This is important because quantifying data in higher resolution produces more accurate results for what’s really on the surface of the Earth…(More)”.

Global Struggle Over AI Surveillance


Report by the National Endowment for Democracy: “From cameras that identify the faces of passersby to algorithms that keep tabs on public sentiment online, artificial intelligence (AI)-powered tools are opening new frontiers in state surveillance around the world. Law enforcement, national security, criminal justice, and border management organizations in every region are relying on these technologies—which use statistical pattern recognition, machine learning, and big data analytics—to monitor citizens.

What are the governance implications of these enhanced surveillance capabilities?

This report explores the challenge of safeguarding democratic principles and processes as AI technologies enable governments to collect, process, and integrate unprecedented quantities of data about the online and offline activities of individual citizens. Three complementary essays examine the spread of AI surveillance systems, their impact, and the transnational struggle to erect guardrails that uphold democratic values.

In the lead essay, Steven Feldstein, a senior fellow at the Carnegie Endowment for International Peace, assesses the global spread of AI surveillance tools and ongoing efforts at the local, national, and multilateral levels to set rules for their design, deployment, and use. It gives particular attention to the dynamics in young or fragile democracies and hybrid regimes, where checks on surveillance powers may be weakened but civil society still has space to investigate and challenge surveillance deployments.

Two case studies provide more granular depictions of how civil society can influence this norm-shaping process: In the first, Eduardo Ferreyra of Argentina’s Asociación por los Derechos Civiles discusses strategies for overcoming common obstacles to research and debate on surveillance systems. In the second, Danilo Krivokapic of Serbia’s SHARE Foundation describes how his organization drew national and global attention to the deployment of Huawei smart cameras in Belgrade…(More)”.

Americans’ Views of Government: Decades of Distrust, Enduring Support for Its Role


Pew Research: “Americans remain deeply distrustful of and dissatisfied with their government. Just 20% say they trust the government in Washington to do the right thing just about always or most of the time – a sentiment that has changed very little since former President George W. Bush’s second term in office.

Chart shows low public trust in federal government has persisted for nearly two decades

The public’s criticisms of the federal government are many and varied. Some are familiar: Just 6% say the phrase “careful with taxpayer money” describes the federal government extremely or very well; another 21% say this describes the government somewhat well. A comparably small share (only 8%) describes the government as being responsive to the needs of ordinary Americans.

The federal government gets mixed ratings for its handling of specific issues. Evaluations are highly positive in some respects, including for responding to natural disasters (70% say the government does a good job of this) and keeping the country safe from terrorism (68%). However, only about a quarter of Americans say the government has done a good job managing the immigration system and helping people get out of poverty (24% each). And the share giving the government a positive rating for strengthening the economy has declined 17 percentage points since 2020, from 54% to 37%.

Yet Americans’ unhappiness with government has long coexisted with their continued support for government having a substantial role in many realms. And when asked how much the federal government does to address the concerns of various groups in the United States, there is a widespread belief that it does too little on issues affecting many of the groups asked about, including middle-income people (69%), those with lower incomes (66%) and retired people (65%)…(More)”.

How can data stop homelessness before it starts?


Article by Andrea Danes and Jessica Chamba: “When homelessness in Maidstone, England, soared by 58% over just five years, the Borough Council sought to shift its focus from crisis response to building early-intervention and prevention capacity. Working with EY teams and our UK technology partner, Xantura, the council created and implemented a data-focused tool — called OneView — that enabled the council to tackle their challenges in a new way.

Specifically, OneView’s predictive analytic and natural language generation capabilities enabled participating agencies in Maidstone to bring together their data to identify residents who were at risk of homelessness, and then to intervene before they were actually living on the street. In the initial pilot year, almost 100 households were prevented from becoming homeless, even as the COVID-19 pandemic took hold and grew. And, overall, the rate of homelessness fell by 40%. 

As evidenced by the Maidstone model, data analytics and predictive modeling will play an indispensable role in enabling us to realize a very big vision — a world in which everyone has a reliable roof over their heads.

Against that backdrop, it’s important to stress that the roadmap for preventing homelessness has to contain components beyond just better avenues for using data. It must also include shrewd approaches for dealing with complex issues such as funding, standards, governance, cultural differences and informed consent to permit the exchange of personal information, among others. Perhaps most importantly, the work needs to be championed by organizational and governmental leaders who believe transformative, systemic change is possible and are committed to achieving it.

Introducing the Smart Safety Net

To move forward, human services organizations need to look beyond modernizing service delivery to transforming it, and to evolve from integration to intuitive design. New technologies provide opportunities to truly rethink and redesign in ways that would have been impossible in the past.

A Smart Safety Net can shape a bold new future for social care. Doing so will require broad, fundamental changes at an organizational level, more collaboration across agencies, data integration and greater care co-ordination. At its heart, a Smart Safety Net entails:

  • A system-wide approach to addressing the needs of each individual and family, including pooled funding that supports coordination so that, for example, users in one program are automatically enrolled in other programs for which they are eligible.
  • Human-centered design that genuinely integrates the recipients of services (patients, clients, customers, etc.), as well as their experiences and insights, into the creation and implementation of policies, systems and services that affect them.
  • Data-driven policy, services, workflows, automation and security to improve processes, save money and facilitate accurate, real-time decision-making, especially to advance the overarching priority of nearly every program and service; that is, early intervention and prevention.
  • Frontline case workers who are supported and empowered to focus on their core purpose. With a lower administrative burden, they are able to invest more time in building relationships with vulnerable constituents and act as “coaches” to improve people’s lives.
  • Outcomes-based commissioning of services, measured against a more holistic wellbeing framework, from an ecosystem of public, private and not-for-profit providers, with government acting as system stewards and service integrators…(More)”.