Phyllis Migwi Shares Microsoft’s AI Strategy

The transformative potential of Artificial Intelligence (AI) stands as a symbol of innovation spanning industries worldwide. In Africa, however, AI remains nascent, and its transformative capabilities are yet to be extensively exploited. There are critical areas that Africa needs to address to optimize the AI ecosystem and realize the full benefits of this emergent technology.

According to a recent report by QUBIT Hub, these critical areas include limited data sets, a shortage of AI talent, limited digital infrastructure, and regulatory challenges. Companies utilizing AI also face concerns regarding privacy, safety, security, inclusivity, accountability, and consumer protection. Microsoft, considered a leader in this field, offers AI solutions for businesses worldwide. In this interview, Phylis Migwi, Microsoft’s Country General Manager in Kenya, shares the global tech company’s vision for AI in Africa.

Will AI take your job? Are you worried about future generations?

Microsoft is focused on developing AI solutions that will augment rather than replace the workforce. We have seen that, when people are freed from repetitive or tedious work, they feel empowered to contribute their best work and tap into their human ingenuity, and everyone wins.

However, as AI systems evolve, we expect that AI advances will change the nature of some jobs and work, and even create new jobs that didn’t exist before. These shifts are similar to the changes we’ve seen with other major technological advances such as the invention of the telephone or the car. For example, before the internet was around, no one could have predicted that there would be jobs such as cybersecurity analyst or social media manager.

We expect this shift will require new ways of thinking about skills and training to ensure that workers are prepared for the future and that there is enough talent available for critical jobs.

Considering the evolving landscape of AI, what is your perspective on the potential benefits and challenges that a future powered by AI may bring?

In a world of increasing economic complexity, new AI tools have the power to revolutionise many types of work and provide substantive productivity gains.

AI is going to be the ultimate amplifier. It will augment the work that people do by freeing up time for more creativity, imagination, and human ingenuity – leading to not only an increase in productivity, but satisfaction.

In line with this, Microsoft has introduced an AI-powered platform – Copilot – that offers a range of capabilities to enhance productivity, creativity and performance.  The very name “Copilot” speaks to the role AI will ultimately play in society, not acting as the pilot or on autopilot, but as an assistant to people – elevating their job functions rather than replacing them. In fact, Microsoft’s research among professionals using Copilot showed that 70 percent of those surveyed felt that they are more productive and another 77 percent saying they wouldn’t want to give their AI assistant up.

We’re are also seeing significant transformation across key areas of society as adoption of AI transforms more organisations. Healthcare nonprofit Amref Health Africa, for example, is working with Microsoft to automate processes that support functions such as programme development and fundraising – leading to stronger donor relationships. The organisation is also working with Microsoft’s AI for Good Lab in Kenya to develop a machine learning model to more accurately predict risk of malnutrition in children.

Kenya, with its up-and-coming tech hubs and dynamic young population, has significant potential to benefit from the global wave of AI transformation. We have a strong track-record of investing in its startups, securing the second highest amount of startup funding in Africa between January 2015 and November 2022, and as the birthplace of M-Pesa, Ushahidi and the iHub, we continue to have a significant impact on Africa’s startup scene.

It would, however, be misguided of us to focus only on the benefits of AI without being clear on the challenges. Ultimately, Microsoft believes that when you create technologies that can change the world, you must also ensure that the technology is used responsibly. We are committed to creating responsible AI by design. Our work is guided by a core set of principles: fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability. We are putting those principles into practice across the company to develop and deploy AI that will have a positive impact on society.

How do you make AI work in an emerging market without superimposing western standards on Africans?

As we navigate this AI-powered future, our journey must be underpinned by responsible and sustainable innovation, ensuring that our progress remains aligned with the human values, societal norms, and the needs that define Africa’s vastly diverse cultures. This involves a deep engagement with the continent’s unique challenges, recognizing their complexity and prioritizing those that AI could help solve. Collaborating with relevant stakeholders will be key to ensuring that AI solutions are not just technologically advanced, but also culturally attuned and genuinely beneficial to African societies.

A thoughtful approach to regulation is also key to ensuring that our exploration of AI’s vast potential is conducted with responsibility and ethical foresight, taking Africa’s unique socioeconomic and cultural context into account. Legislation should consider issues such as protecting data privacy, ethical AI use, and minimising bias with the utmost care while simultaneously nurturing the opportunity this revolutionary technology presents.

Given the varied stages of AI adoption and regulations across the continent’s 54 countries, a cooperative and thoughtful approach is needed to standardize data and address these challenges effectively. As AI usage expands, more African countries are expected to introduce regulations to guide its development and deployment responsibly.

So far, what key challenges do organizations in Africa typically face when adopting AI, and how can they overcome these challenges to fully harness the benefits of AI technology?

In a broad sense, the primary barriers to AI adoption in Africa are related to lack of infrastructure, connectivity, skills and essential digital tools. Even where people are able to access AI-based technologies, they are typically hindered by the amount, quality and representativeness of the data available for use in AI.

When it comes to organizations that are already on their AI journeys, the biggest challenges typically come down to demystifying the impact on personnel.

In a region where unemployment rates are high and true understanding of AI is still a work in progress, employees have an inherent reservation towards scaled adoption of AI and significant amounts of education and change management will be required. Employees need to understand that AI is predominantly an enabling capability that enhances productivity and reduces the amount of time spent on menial tasks, in order for them to embrace the application of AI and realise the benefits.

Organizations in Africa are encouraged to develop strong AI governance systems. Can you share insights into how companies can establish guiding principles, especially focusing on fairness, reliability, and inclusiveness?

Using key principles as a foundation, organizations can then develop an internal governance system to help them engage with AI responsibly and share resources, guidance, and tools with their engineering teams.

To establish fairness, for example, companies need to understand the limitations of AI predictions and recommendations. Though AI can provide helpful suggestions, at the end of the day final decisions must be made by an accountable person.

In the same way, the developers designing and building these AI systems need to understand how bias might affect the final solution. They can then mitigate that bias by using diverse datasets to train AI models so they can learn and evolve without developing prejudices.

For example, consider an AI system designed to help employers screen job applicants. When trained on data from public employment records, this system might “learn” that most software developers are male and favour men over women when recommending candidates. To help them develop AI systems that treat everyone fairly, developers can leverage tools, methodologies, techniques, and other resources that help detect and mitigate biases.

It’s also essential that AI systems operate reliably, safely and consistently, not just under normal circumstances but during unexpected situations too. How AI ultimately behaves is generally determined by the range of circumstances developers anticipate during design and testing. It’s therefore critical for developers to ensure AI can handle even unanticipated situations by employing rigorous testing.

We believe that rigorous testing is essential during system development and deployment to ensure AI systems can respond safely in unanticipated situations and edge cases, don’t have unexpected performance failures, and don’t evolve in ways that are inconsistent with original expectations.

After testing and deployment, it’s equally important that organizations properly operate, maintain, and protect their AI systems over the lifespan of their use. Otherwise, AI systems can become unreliable or inaccurate if they are not maintained properly. Ultimately, because AI should augment and amplify human capabilities, people need to play a critical role in making decisions about how and when an AI system is deployed, and whether it’s appropriate to continue to use it over time. Human judgment will be key to identifying potential blind spots and biases in AI systems.

Finally for AI to be inclusive, it must incorporate and address a broad range of human needs and experiences. AI has huge potential to improve access to a wide range of essential services such as education and healthcare. But in order to realise this potential, developers need to adopt inclusive design practices, whereby they address aspects of the product environment that could unintentionally exclude people.

Aside from using the Inclusive Design Toolkit, developers should consider leveraging commonly used accessibility standards and have people with disabilities test their systems.

Bias in AI systems is a significant concern. How does Microsoft address bias in AI development, particularly in terms of diverse datasets and training models to mitigate prejudices?

Microsoft is proactively focused on the risks created by data bias and conducts research and develops tools to make detecting and correcting data bias more feasible in complex AI systems.

We’ve introduced tools including the InterpretML toolkit, which enables developers and data scientists to understand model behaviour and provide model explanations to business stakeholders and customers, and Fairlearn, which allows developers and data scientists to leverage specialised algorithms to ensure fairer outcomes for everyone.  It is a transformative time for AI and we need to have wide-ranging and deep conversations and commit to joint action to define the guardrails for the future.

We believe we should focus on three key goals: Ensure that AI is built and used responsibly and ethically, that AI advances international competitiveness and national security, and that AI serves society broadly, not narrowly.

There is the issue of data quality in Africa, how can technology, including AI itself, be leveraged to enhance data quality in Africa, and are there specific tools or approaches that organizations are adopting for this purpose?

A crucial step towards countries effectively developing, deploying, and managing AI technologies and paving the way for technological advancement and economic growth relies on the adoption of cloud-first policies that can provide the scalable, cost-effective infrastructure essential for handling AI’s computational demands and large data sets. Cloud computing democratizes access to advanced AI technologies, while fostering an environment of innovation and collaboration. And generally, the companies most advanced in using AI report use of hybrid architectures with on-premise and cloud based storage, while the less advanced predominantly rely on on-premise platforms.

With a view to providing more organizations with the capacity to drive progress, Microsoft has invested in datacenters on the continent, in Cape Town and Johannesburg, as well as in five edge sites and subsea cables and gateway colocation deployments.

AI and machine learning are also helping to improve availability and quality of data for companies like M-KOPA, a fintech company that serves the financially excluded. While the idea is simple, advances in AI and machine learning on Azure have improved the business model for M-KOPA. The use of AI has helped M-KOPA achieve significant increases in customer repayment performance – particularly for the follow-on products and services that M-KOPA offers to customers once they have successfully repaid their initial loan.

Tracking the health of its lending portfolio is an absolute bedrock competency of the organization. As such, the company uses machine learning models to forecast the repayment of every single loan in its portfolio – every single day. And that gives M-KOPA very early signals on any changes in repayment patterns, which in turn helps the company strike an important balance between inclusive pricing, financial responsibility and, ultimately, customer protection.

Microsoft has expanded its AI for Good Lab into Africa and established the Africa AI Innovation Council. Could you elaborate on Microsoft’s initiatives and their objectives in driving AI skills and capacity in Africa?

We have invested heavily in skilling and capacity-building to catapult African digital economies into the future – more than 4 million young people across Africa have been upskilled over the last five years through various skilling and employability programs, with a further commitment to train and certify at least 1 million women across Africa by June 2025.

Microsoft’s Africa Development Centre (ADC), our first-ever engineering offices on the continent, operates from three different locations in Africa, including Nairobi, Kenya; Lagos, Nigeria and Cairo, Egypt. Our engineers are fueling AI and machine learning innovation across numerous domains from healthcare, agriculture, finance, and human centric automation.

The ADC also continues to play a pivotal role in growing and enhancing the African tech talent landscape through various ecosystem engagements such as university faculty skilling programs, university curriculum reviews and student hackathons.

For organizations, we also have a host of learning platforms available. Microsoft’s AI Business School, for example, offers an online leadership series that covers how to drive business impact by creating an effective AI strategy, enabling an AI-ready culture, innovating responsibly, and more.

As AI becomes more pervasive, how does Microsoft ensure that AI systems comply with privacy laws, and what measures are taken to secure personal and business information?

We create our products with security and privacy incorporated through all phases of design and implementation. We provide tools and clear choices so people can control their data, including through tools to access, manage, and delete personal data and stored conversation history.

Our approach to privacy in AI systems is grounded in our longstanding belief that privacy is a fundamental human right. We are committed to continued compliance with all applicable laws, including privacy and data protection regulations, and we support accelerating the development of appropriate guardrails to build trust in AI systems.

We believe the approach we have taken to enhance privacy in our AI technology will help provide clarity to people about how they can control and protect their data in our new generative AI products.

Bing, for example, has processes in place to address personal information appearing in search features today, which will flow through to new Bing features.

How is Microsoft contributing to the development of local AI skills, and what initiatives are in place to support the growth of AI talent in Africa?

Recognizing the need to nurture the next generation of African innovators, we are investing in training and resources to develop digital skills and AI literacy. These efforts are encapsulated in Microsoft’s wide range of skilling and employability programs. As mentioned above, more than 4 million young people have been upskilled in the last five years alone. One example of the kind of programs we invest in is our recent AI Skills Initiative, which offers new and free courses for workers to learn about how to harness the power of AI and drive Africa’s growth in the digital age.

Beyond guiding principles, how does Microsoft ensure ethical considerations are integrated into AI development, especially in regions with diverse cultural and ethical perspectives?

As mentioned above, inclusive design practices can help system developers understand and address potential barriers in a product environment that could unintentionally exclude people. By addressing these barriers, we create opportunities to innovate and design better experiences that benefit everyone. Inclusive Design is a methodology, born out of digital environments, that enables and draws on the full range of human diversity. Most importantly, this means including and learning from people with a range of perspectives.

How does Microsoft collaborate with local tech ecosystems and startups in Africa to foster innovation, share expertise, and accelerate the adoption of AI technologies?

Through investments made over the past three decades, we have been committed to closing the digital skills gap and empowering businesses and individuals in Africa with the tools needed to solve key challenges using technology. Key investments have included:

• Broadscale skilling initiatives such as the Youth in Africa initiative, which aims to upskill 50 million youth and create 25 million jobs across Africa and in Kenya, in partnership with the African Development Bank.
• Partnerships with the Ministry of Education in Kenya on various impact-led programs that include the distribution of learning materials to 16 million students, and training of over 350,000 teachers.

Our capacity building efforts have extended to growing engineering and developer talent through the Microsoft Africa Development Center. The ADC will play an important part in growing and enhancing African tech talent.

We also see SMEs as important drivers of economic growth and have also committed to supporting SMEs through a number of initiatives. We believe they have a critical role in accelerating economic development and serving unmet needs of African markets.

Already, we are witnessing the AI revolution firsthand through our collaborative efforts with government, startups and organizations across the continent to reimagine services, expand businesses, and instil innovation across society.

However, beyond investment in essential infrastructure, digital assets and capabilities, we believe the private sector also has a key role in working with the public sector to help shape AI regulation.

Microsoft was one of the first major technology companies to call for thoughtful government regulation on facial recognition technology and is committed to creating responsible AI by design through our Responsible AI standard.

There is a rich and active global dialogue about how to create principled and actionable norms to ensure organizations develop and deploy AI responsibly – and this conversation is happening at all levels across organizations.

For our part, Microsoft continues to engage with the African Union (AU) and national governments in priority markets to establish Microsoft as a trusted voice on responsible AI in Africa.  This includes advancing a regulatory architecture for AI using the Governing AI Blueprint and contributing to the development of new AI voluntary codes, laws, and regulations through targeted campaigns, the drafting of AI papers and launch of skilling initiatives.

What do you think about Google’s Bard?

We believe AI is a defining technology of our time, and we are optimistic about what it can and will do for people, industry, and society. While saying that, responsible AI is a journey, and as we are making significant progress, we expect to encounter challenges that require us to pause, reflect, and adjust.