The building blocks of Microsoft’s responsible AI program | Microsoft On The Issues

The building blocks of Microsoft’s responsible AI program | Microsoft On The Issues

The pace at which artificial intelligence (AI) is advancing is remarkable. As we look out at the next few years for this field, one thing is clear: AI will be celebrated for its benefits but also scrutinized and, to some degree, feared. It remains our belief that, for AI to benefit everyone, it must be developed and used in ways which warrant people’s trust. Microsoft’s approach, which is based on our AI principles, is focused on proactively establishing guardrails for AI systems so that we can make sure that their risks are anticipated and mitigated, and their benefits are maximized…

Governance as a foundation for compliance

While there is much that is new and unchartered in the domain of responsible AI, there’s also much that can be learned from adjacent domains. Our responsible AI governance approach borrows the hub-and-spoke model that has worked successfully to integrate privacy, security and accessibility into our products and services…

Developing rules to enact our principles

In the fall of 2019, we published internally the first version of our Responsible AI Standard, a set of rules for how we enact our responsible AI principles underpinned by Microsoft’s corporate policy. We published the first version of the Standard with an eye to learning, and with a humble recognition that we were at the beginning of our effort to systematically move from principles to practices. Through a phased pilot across 10 engineering groups and two customer-facing teams, we learned what worked and what did not. Our pilot teams appreciated the examples of how responsible AI concerns can arise. They also struggled sometimes with the open-endedness of the considerations laid out in the Standard and expressed a desire for more concrete requirements and criteria. There was a thirst for more tools, templates, and systems, and for a closer integration with existing development practices…

Drawing red lines and working through the grey areas

In the fast-moving and nuanced practice of responsible AI, it is impossible to reduce all the complex sociotechnical considerations into an exhaustive set of pre-defined rules. This led us to create a process for ongoing review and oversight of high-impact cases and rising issues and questions…

Evolving our mindset and asking hard questions

Today, we understand that it is critically important for our employees to think holistically about the AI systems we choose to build. As part of this, we all need to think deeply about and account for sociotechnical impacts. That’s why we’ve developed training and practices to help our teams build the muscle of asking ground-zero questions, such as, “Why are we building this AI system?” and, “Is the AI technology at the core of this system ready for this application?”

In 2020, our mandatory Introduction to Responsible AI training helped more than 145,000 employees learn the sensitive use process, the Responsible AI Standard and the foundations of our AI principles…

Pioneering new engineering practices

Privacy, and the GDPR experience in particular, taught us the importance of engineered systems and tools for enacting a new initiative at scale and ensuring that key considerations are baked in by design.

As we have been rolling out our responsible AI program across the company, the existence of engineering systems and tools to help deliver on our responsible AI commitments has been a priority for our teams. Although tooling – particularly in its most technical sense – is not capable of the deep, human-centered thinking work that needs to be undertaken while conceiving AI systems, we think it is important to develop repeatable tools, patterns and practices where possible so the creative thought of our engineering teams can be directed toward the most novel and unique challenges, not reinventing the wheel. Integrated systems and tools also help drive consistency and ensure that responsible AI is part of the everyday way in which our engineering teams work…

Scaling our efforts to develop AI responsibly

As we look ahead, we’ll focus on three things: first, consistently and systematically enacting our principles through the continued rollout of our Responsible AI Standard; second, advancing the state of the art of responsible AI through research-to-practice incubations and new engineering systems and tools; third, continuing to build a culture of responsible AI across the company.

We are acutely aware that, as the adoption of AI technologies accelerates, new and complex ethical challenges will arise. While we recognize that we don’t have all the answers, the building blocks of our approach to responsible AI at Microsoft are designed to help us stay ahead of these challenges and enact a deliberate and principled approach. We will continue to share what we learn, and we welcome opportunities to learn with others.

The rest of the post The building blocks of Microsoft’s responsible AI program appeared first on Microsoft On the Issues.

from Microsoft On the Issues

Microsoft and OpenAI partner to propose digital transformation of export controls | Microsoft On The Issues

Microsoft and OpenAI partner to propose digital transformation of export controls | Microsoft On The Issues

The worlds of technology, trade and national security policy are converging as never before. Governments and non-government actors are vigorously debating how to ensure that powerful technologies are used by trustworthy actors and to good ends – and are increasingly looking to export controls as a way to achieve this. Targeted export controls on end uses and users of concern are needed to protect national security interests on the one hand without provoking serious unintended consequences on the other. However, these can be difficult to administer and enforce. The time is right for a digital transformation of export controls – a new approach that leverages novel digital solutions within sensitive and important technology itself to better protect it from uses that harm national security, while preserving its beneficial uses. Microsoft and OpenAI have joined together to work on these solutions. In a submission to the U.S. government yesterday and here, we describe how a digitally transformed export controls system would work and the substantial benefits it would provide.1

The challenge

Following a mandate in the Export Control Reform Act of 2018 (ECRA), the U.S. Department of Commerce’s Bureau of Industry and Security (BIS) has undertaken efforts to identify and control the export of “emerging” or “foundational” technologies essential to U.S. national security. In comment periods ending in January 2019 and yesterday, BIS sought help from industry and others on how to identify and approach control of these emerging and foundational technologies…

The solution: How a digital transformation of export controls will work

The government will set policies that determine who can access sensitive technologies and for what purpose from an export controls and national security perspective. These policies would then be implemented and enforced within the protected technology itself, as well as hardening the infrastructure around it to prevent circumvention. These solutions can protect against problematic users and uses in a more targeted, effective and dynamic way – not just at initial access but continuously in a deployed environment…

Applications beyond export controls

…Employed appropriately, these digital solutions will provide valuable commercial benefits to users, as well as a far more powerful, dynamic and targeted method for controlling exports of important technologies. We look forward to continuing to exchange ideas with a range of stakeholders interested in pursuing such solutions.

The rest of the post Microsoft and OpenAI partner to propose digital transformation of export controls appeared first on Microsoft on the Issues.

from Microsoft on the Issues

1 For our full U.S. government submission, see Microsoft and OpenAI Comment on Advance Notice of Proposed Rulemaking Regarding Review of Controls for Certain Foundational Technologies.

Open Data Campaign: Exploring the power of open data | Microsoft On The Issues

Open Data Campaign: Exploring the power of open data | Microsoft On The Issues

In April, we announced the launch of the Open Data Campaign to close the “data divide” and ensure that organizations of all sizes have access to the data they need to innovate with artificial intelligence (AI). To demonstrate the importance of being more open with data and the need to share data to address pressing issues, we committed to the development of 20 data collaborations by 2022. Through these collaborations, we will work with partners to address issues that are “top of mind” and require urgent action. One thing remains true in these uncertain times: To tackle the pressing societal issues we face today – everything from climate change to COVID-19, justice reform to digital access – people and organizations need access to the data that can help unlock the power of innovation and technology.

The past several months have accelerated this work in many ways, and we are learning a lot. The campaign launched just as the world was grappling with the COVID-19 pandemic, and the value of being more open with data become clear in new, undeniable ways. For example, The Alan Turing Institute had been leading an air quality project, with the support of Microsoft’s AI for Earth program, collecting data from across London to understand air pollution. As the summer progressed, however, the institute discovered that analysis of these same data streams could also be used to understand London’s “busyness” as COVID-19 restrictions were eased. It’s just one example of how open data and data collaboration can provide valuable insights beyond their initial focus.

When armed with the right data, organizations are empowered to make decisions that positively impact their employees, customers and the communities they serve. In many ways, data is taking center stage in the response from governments and companies to fuel promising solutions and ideas.

Since launching the campaign, we have moved forward in forging strong data collaborations with partners – focusing our efforts on issues that require commitment, collaboration and urgent action to address:

  • Climate change: On September 1, Microsoft joined with Allianz, Amazon and S&P Global to announce plans to launch the Climate Finance Foundation, led by the Linux Foundation. Climate change is one of the most pressing challenges of the 21st century, and we are committed to being a part of the solution. The ability to leverage high-quality, open, corporate sustainability data will be critical for enabling the investment community to make informed decisions based on accurate and reliable economic models around corporate climate-related risk and opportunity. Microsoft is investing heavily in sustainability, and we have made a commitment to share relevant sustainability data to the open Data Commons supported by this effort.
  • COVID-19: The Alan Turing Institute partnered with the Greater London Authority, supported by Microsoft and the London Data Commission, to demonstrate the value of data sharing to help support London’s response and recovery to COVID-19. This pilot is looks at London’s “busyness” – or movement around the city – through multiple data sources as COVID-19 restrictions are relaxed to monitor how people are responding to the changes. Microsoft provided Azure AI and cloud infrastructure and services to support the COVID-19 pilot. Transport for London is already operationalizing the outputs from this pilot.
  • Digital access and education: When we launched the Open Data Campaign in April, one of our first announcements was that we would be working with the Open Data Institute and BroadbandNow to help address the issue of broadband availability. With COVID-19 taking a significant toll on students’ ability to access face-to-face education, we know that this need is more urgent than ever. As governments, policymakers, nonprofits, and organizations around the world are looking at ways to target resources that serve students more effectively, we are excited to share that we’ll be launching an Open Data Challenge in October to look at the impact of digital access and COVID-19 on young students’ education. We look forward to sharing more details on this collaboration in the coming weeks…

The rest of the post Open Data Campaign: Exploring the power of open data appeared first on Microsoft on the Issues.

from Microsoft on the Issues

Mobilizing AI for Health to fight against COVID-19 | Microsoft On The Issues

Mobilizing AI for Health to fight against COVID-19 | Microsoft On The Issues

On January 29, 2020, we announced the launch of AI for Health, an initiative to advance the health of people and communities around the world. This five-year commitment was created to empower nonprofits, researchers and organizations with AI and data science tools.

Since then, the world has changed. As of the time of writing, the COVID-19 virus has infected more than 1.4 million people around the world. The crisis has made it painfully clear that health transcends every border, impacting every person on the planet.

This is part of Microsoft’s larger commitment toward fighting COVID-19, as we are working to support remote education and empower students around the world, enabling businesses to work from home, securing needed medical supplies and supporting local communities. We hope this added commitment empowers researchers and organizations to solve this crisis…

AI for Health COVID-19 focus areas

Given the global scale of the pandemic, technology will play a critical role in nearly every facet of addressing COVID-19, from using AI to crunch massive datasets to analyzing disease vectors and identifying treatment impacts. We will collaborate with nonprofits, governments and academic researchers on solutions, and bring our experience to the table, providing access to Microsoft AI, technical experts, data scientists and other resources…

Understanding and tracking our progress against COVID-19

We want the world to better understand COVID-19. As such, we have developed a set of interactive visualizations, available below, so everyone has full transparency into the scope of the problem and the progress we are making together to heal the world. We will continue to update and refine this visualization with new data and insights…

COVID-19 is a global problem and finding a solution will take all our efforts. We are humbled and honored to work with researchers across the globe and support them with this additional dedicated support from AI for Health.

The rest of the post Mobilizing AI for Health to fight against COVID-19 appeared first on Microsoft on the Issues.

from Microsoft on the Issues via IFTTT

Using AI to advance the health of people and communities around the world | Microsoft on the Issues

Using AI to advance the health of people and communities around the world | Microsoft on the Issues

The health of people and communities around the world has been improving over time. For example, the steep decline in child and maternal mortality is a key indicator of positive momentum.

However, progress has not been equal across the globe, and there is a great need to focus on societal issues such as reducing health inequity and improving access to care for underserved populations. While researchers work to unlock life-saving discoveries and develop new approaches to pressing health issues, advancements in technology can help accelerate and scale new solutions.

That is why we are launching AI for Health, a new $40 million, five-year program to empower researchers and organizations with AI to improve the health of people and communities around the world. The program is underpinned with a strong foundation of privacy, security and ethics, and was developed in collaboration with leading health experts who are driving important medical initiatives. AI for Health is the fifth Microsoft AI for Good program, a $165 million initiative to empower researchers, nonprofits and organizations with advanced technologies to help unlock solutions to the biggest challenges facing society today.

The AI for Health initiative will focus on three key areas:

  • Quest for discovery. Accelerating medical research to advance prevention, diagnoses and treatment of diseases
  • Global health insights. Increasing our shared understanding of mortality and longevity to protect against global health crises
  • Health equity. Reducing health inequity and improving access to care for underserved populations

AI for Health is a philanthropic initiative that complements our broader work in Microsoft Healthcare. Through AI for Health, we will support specific nonprofits and academic collaboration with Microsoft’s leading data scientists, access to best-in-class AI tools and cloud computing, and select cash grants.

I am honored to lead AI for Health as part of my mission at Microsoft to fuse AI and data to address the world’s greatest challenges. As a tech company, it is our responsibility to ensure that organizations working on the most pressing societal issues have access to our latest AI technology and the expertise of our technical talent.

The rest of this fascinating post Using AI to advance the health of people and communities around the world appeared first on Microsoft on the Issues.

from Microsoft on the Issues

Six principles to guide Microsoft’s facial recognition work

In his recent speech at the Brookings Institution, Brad Smith talked about the urgent need for governments to adopt laws to regulate facial recognition technology. The recommendations, outlined in an accompanying blog post, frame a broader journey we as a society must take to address important questions about the technology while it is still in its infancy before it’s too late to put the facial recognition genie back in its bottle. He also introduced the principles that will guide Microsoft in how we develop and deploy facial recognition technology. The principles are:


  1. Fairness. We will work to develop and deploy facial recognition technology in a manner that strives to treat all people fairly.
  2. Transparency. We will document and clearly communicate the capabilities and limitations of facial recognition technology.
  3. Accountability. We will encourage and help our customers to deploy facial recognition technology in a manner that ensures an appropriate level of human control for uses that may affect people in consequential ways.
  4. Non-discrimination. We will prohibit in our terms of service the use of facial recognition technology to engage in unlawful discrimination.
  5. Notice and consent. We will encourage private sector customers to provide notice and secure consent for the deployment of facial recognition technology.
  6. Lawful surveillance. We will advocate for safeguards for people’s democratic freedoms in law enforcement surveillance scenarios and will not deploy facial recognition technology in scenarios that we believe will put these freedoms at risk.


We explain these principles in more detail here. Our goal is to make these principles operational by the end of March 2019. But even as we implement them, we do so knowing that the issues are novel and complex and that we still have much to learn. We fully anticipate the principles will evolve over time based on our experience, the experience of others, and the ongoing conversations we will have around facial recognition technology with customers, public officials, technologists, academics, civil society groups, and multi-stakeholder organizations such as the Partnership on AI.  We remain committed to continuing these critical conversations, to advocating for laws that keep pace with the inevitable advances of this technology, and to sharing what we learn.


The post Six principles to guide Microsoft’s facial recognition work appeared first on Microsoft on the Issues.

Announcing the first AI for Accessibility grantee: Zyrobotics


Satya Nadella (C) with Zyrobotics CEO Dr. Johnetta MacCalla (L) and CTO Dr. Ayanna Howard (R) courtesy of Microsoft

Announcing the first AI for Accessibility grantee: Zyrobotics

via Tumblr published on October 06, 2018 at 02:33AM