As we mark Safer Internet Day 2026, we’re reflecting on a simple but enduring principle: safety must be designed into online services, not bolted on. Microsoft’s work in this space spans more than two decades—from technology solutions like PhotoDNA to our investments in responsible gaming, public-private partnerships, and empowering users through education. This foundation guides our approach as we help individuals and families navigate a rapidly evolving landscape shaped by new technologies and new risks and as we innovate with next-generation AI offerings. At a moment when 91% of people tell us they worry about harms introduced by AI, our commitment to responsible innovation has never been more important—especially for our youngest users.

Read on for more about our longstanding efforts to create a safer digital environment, plus key findings from our Global Online Safety Survey and new examples of our work to empower families and communities through tools, research, and educational resources—including the latest release in Minecraft Education’s CyberSafe series.  

Ten years of safety research 

2026 marks the tenth year of our annual Global Online Safety Survey research. For a decade, we have invested in surveying teens and adults around the world about their experiences and perceptions of life online—aiming to provide fresh insights to support our collective work. That’s 130,000+ interviews across 37 countries, with the results available on our website. Ten years later, respondents tell us that they feel more connected and more productive, but less safe online.  

This year’s Global Online Safety Survey also highlights the complexity of the digital environment young people now inhabit. Teens’ exposure to risk rose again, with hate speech (35%), scams (29%), and cyberbullying (23%) among the most commonly experienced harms. At the same time, teens demonstrated striking resilience: 72% talked to someone after experiencing a risk, and reporting behavior increased for the second consecutive year. But worries about the misuse of AI continue, underscoring again why safety-by-design for AI is essential, not optional. Find the full results and country-level summaries here. 

Year on year, the research has told a story of evolving online safety risks and of the real-world impact. In 2026, the call to action is more urgent than ever—unless industry can deliver safe and age-appropriate experiences, young people risk losing access to technology. At Microsoft, spanning across our teams from Windows to Xbox, we have sought to continuously evolve our approach and to lead industry in advancing tailored and thoughtful safety solutions.   

Evolving to meet the moment 

Looking ahead, we know we need to continue to build strong guardrails to tackle acute risks and to leverage our experience while being informed by new research, new perspectives, and new technologies. The application process closed yesterday for our first AI Futures Youth Council, to be comprised of teens from across the US and EU. We’re looking forward to bringing those teens together soon for a first meeting to get their direct feedback on the role they want emerging technology to play in their lives and how we can best support their safety.  

Microsoft has partnered with Cyberlite on a second youth-centered initiative to understand how teens aged 13–17 are engaging with AI companions. Through codesign workshops with students in India and Singapore, we’re capturing young people’s own perspectives on the benefits, risks, and emotional dimensions of AI use—insights that will directly inform educational resources for teens, parents, and educators. Early findings from the first workshop in December 2025 show that young people value AI as a judgment free space while also recognizing the tradeoffs: privacy risks, overreliance, and erosion of critical thinking loom larger for them than bad advice.  

We’re also thinking about how we define safety in the next era of Windows, leveraging the Family Safety controls that have been integrated for over a decade. As many countries have raised the local age for digital consent, more parents will have the option to enable parental controls for teens up to the age of 18—leveraging these tools as part of a holistic approach to digital parenting. And to help parents set up and understand Family Safety, we’ve developed a short new guide. 

Safety is also about transparency, empowerment, and education. At Xbox, bringing the joy of gaming to everyone means remaining transparent about the many ways we innovate so players, parents, and caregivers can feel confident that Xbox continues to be a place for positive play. You can read more about our recently published Xbox Transparency Report and the tools and resources available to players on the Xbox Wire blog.  

We’re also excited to announce the latest release in Minecraft Education’s CyberSafe series: CyberSafe: Bad Connection? This series of immersive Minecraft worlds and educational resources is free and helps translate complex risks into fun learning experiences that meet young people in their favorite blocky world. Bad Connection?—the fifth in the series—reflects our commitment to evolving to meet new and challenging risks, with a focus on tackling serious risks related to online recruitment and radicalization. Learn more about how to access this new Minecraft world here.  

The CyberSafe series has reached more than 80 million downloads since 2022 through a partnership between Minecraft Education, Xbox, and Microsoft, helping a generation of young players build the agency, resilience, and digital citizenship they need to navigate an increasingly online world. As part of our commitment to ensure people have the knowledge and skills they need to benefit from technology and stay safe, Microsoft Elevate is empowering educators and students with tools and guidance to build safer, more responsible digital habits, recognizing that AI is transforming how people learn, work, and connect. Our commitment to helping young people access technology safely is also why we’ve partnered with organizations, like the National 4-H Council to prepare young people for an AI-powered world through AI literacy and digital safety curriculum and game-based learning with Minecraft Education. 

As we look ahead, our goal is clear: build technology that is safe by design, guided by evidence, and informed through partnership. The internet has changed profoundly over the past decade, and so too have the expectations of the people who use it. Safer Internet Day is a reminder that progress requires sustained collaboration across industry, civil society, researchers, and families.

—  

Global Online Safety Survey Methodology  

Microsoft has published annual research since 2016 that surveys how people of varying ages use and view online technology. This latest consumer-based report is based on a survey of nearly 15,000 teens (13–17) and adults that was conducted this past summer in 15 countries examining people’s attitudes and perceptions about online safety tools and interactions. Responses to online safety differ depending on the country. Full results can be accessed here.  

 

The post Building a safer digital future, together appeared first on Microsoft On the Issues.

Today, I had the pleasure of joining a range of leaders for timely, impactful discussions on child well-being in the age of AI at the Vatican, building on thoughtful conversations held during the United Nations General Assembly. These issues are top of mind globally, from parents to policymakers to physicians.

At Microsoft, we remain focused on our goal of empowering young people to use technology safely, mindfully, and in pursuit of social, educational, and economic opportunities. That means taking new steps spurred by regulation, such as new age verification measures for our UK Xbox users, as well as adapting our longstanding commitments to responsible AI and child online safety and privacy to build trust in the AI era. Today, we’re sharing new research on youth perspectives, announcing the AI Futures Youth Council to amplify teen voices, and offering policy recommendations to help families navigate the digital world with confidence.

Centering young people’s voices: Announcing the AI Futures Youth Council and new age assurance research

In 2017, Microsoft led the industry with our first Council for Digital Good—a forum where we could hear directly from young people about their experiences and perceptions of online risk. In 2025, with AI reshaping our world—and their future—we again need to center the voices of young people as we think about responsible design for AI and how we set students up for the future. We are actively working with teens from the Asia-Pacific region to develop our first “for teens, by teens” guide to AI chatbots. Today, I’m pleased to announce the upcoming launch of our first “AI Futures Youth Council,” bringing together teens from the US and Europe to have their say on their future. We’ll share more about the application process soon.

We know that a critical precursor to providing young people positive and productive online experiences is understanding which users are young people. Around the globe, the debate over how to achieve age assurance online continues unabated. We have been grateful to work with CIPL and the WeProtect Global Alliance over the last year to explore how to achieve improved age assurance that is consistent with fundamental rights of privacy and access to information. As with any other safety intervention, our goal is to be proportionate and thoughtful where we take new steps, which is why we have focused on gaming in the first instance—reflecting the responsibilities we have to our youngest users and our ongoing commitment to player safety.

To inform our strategy and the broader policy conversation, we partnered with Praesidio Safeguarding to better understand youth perspectives on age assurance approaches across the UK, Ghana, and Indonesia. We are pleased to share that research today. The findings reinforce the importance of transparency, choice, and trust: teens want clear explanations of how their data is used, express concerns about exclusion where formal proof of age is lacking and show varying comfort levels with the use of biometric and behavioral data. Notably, young people value parental involvement but also highlight the need for independence and privacy as they mature. The results also highlight some of the important differences across geographies. For example, teenagers in Ghana often not only share devices with their families but may also share an account—underscoring a need for nuanced global approaches at multiple layers of the technology stack.

These insights underscore our belief that proportionality—matching safeguards to actual risks—is essential to building trust and empowering youth online. They also highlight the need for age assurance models that are inclusive, flexible, and respectful of youth autonomy—especially in global contexts where device and account sharing are common. We remain committed to ongoing dialogue and innovation, ensuring that our solutions evolve alongside the needs and expectations of children, families, and society at large.

Our policy recommendations: Empower young people to use technology safely

We believe technology should empower young people, not put them at risk. Given the diverse range of online services, it is important to remember there is no single “digital seatbelt” to protect and empower young people online.

We therefore offer the following recommendations as policymakers, regulators, and experts continue to discuss these issues, building on our 2024 blog:

• Avoid blanket access restrictions. Age assurance requirements that block full access to a service—except in limited cases like sites dedicated to age-restricted content (e.g., pornography)—can unintentionally limit child rights, such as access to information. Instead, age assurance should be applied at the service level, target specific design features that pose heightened risks, and enable tailored experiences for children.
• Focus on the highest risks for impact, such as content and features associated with documented harms to children, and as determined through democratic processes. Providers should take steps to assess and mitigate risks to children on their services, while ensuring documentation requirements or compliance obligations do not inadvertently undermine safety. A risk-based and proportionate approach—grounded in clear criteria and supported by interoperable standards—can also help ensure that age assurance is applied where most needed, without introducing unnecessary friction. Providers of high-risk services should bear the responsibility of age assurance.

• Strengthen safeguards for AI companions. Recent tragic events have highlighted the need for continued care in developing AI companions, especially where these may be used by young people. At Microsoft, we are building AI services for empowerment and want the right guardrails in place to protect all users but welcome new, commonsense measures such as those enacted in California and Australia to reduce the potential harms related to suicide and self-injury risks, as well as to sexualized or violent content. We will continue to work closely with researchers and experts to understand and mitigate potential risks to young people in this fast-evolving field.
• Incentivize age-appropriate design. Banning kids from online services isn’t the answer, but what constitutes an “age-appropriate” experience will vary. We have supported a duty of care approach to child safety where the duty can be implemented flexibly, guided by thoughtful and evidence-based regulatory guidance. Ongoing research and expert engagement are needed to understand how to advance child safety and rights on diverse services—not just social media.
• Protect the privacy and security of all users. Tailoring age assurance requirements will help enable proportionate approaches to data processing. Current proposals for age verification by app stores risk creating significant privacy risks by collecting sensitive information and sharing unnecessary age data with a wide variety of services while also not solving the challenges lawmakers want to address. We continue to support federal privacy legislation in the US and encourage global efforts to develop standards and certifications for age assurance providers. Trusted credential sharing can also increasingly be enabled by emerging digital identity ecosystems—including government-issued IDs and wallet-based models—that preserve mutual privacy between issuers and relying parties.
• Support, not overwhelm Our Global Online Safety Survey results show that while parents might underestimate the risks teens face online, teens are most likely to turn to a parent for help. Parents should not face a deluge of notifications nor bear the sole responsibility for safety but have access, awareness, and education on family safety tools that can help them make informed choices appropriate for their family and their values.
• Foster multistakeholder collaboration. We believe it’s essential to elevate the voices and perspectives of young people, as well as for regulators and industry to engage with civil society and partner to advance practical solutions. As child safety regulations come into force, it will also be important to get feedback from affected communities on where regulation may have adverse rights impacts, as well as to understand where harm may have been averted. Public education will be needed to help all users understand why their online experiences might be changing.

We will continue learning, listening, and collaborating, especially with our new Council, and look forward to sharing our insights.

 

 

The post Uplifting and empowering young people for an AI future appeared first on Microsoft On the Issues.

As Abu Dhabi and Dubai kick off a significant week hosting annual energy and technology conferences, we want to share details of our ongoing and planned investments in the United Arab Emirates. Roughly two and a half years ago, we embarked on a new AI initiative with the encouragement and support of both the United States and UAE governments. Much of this progress has involved a new partnership with G42, the UAE’s sovereign AI company, with whom we’re making critical progress.   

All told, Microsoft will invest $15.2 billion USD in the UAE between the start of this initiative in 2023 and the end of this decade, in 2029. This is not money raised in the UAE. It’s money we’re spending in the UAE. And as we do everywhere in the world, we’re focused not just on growing our business but also on contributing to the local economy. This involves bringing together three critical factors – technology, talent, and trust.   

On some days, it feels like the tech sector is gripped in a rhetorical race to announce ever larger, sky-high numbers. We believe in moving fast while staying grounded and being transparent about our investment details. And we want to share our strong conviction that our investments benefit the shareholders of our company, the people of the UAE, and the relationship between our two nations.   

Our investment numbers  

Our $15.2 billion USD investment includes the following:  

• Beginning in 2023 and through the end of this calendar year, Microsoft will have invested and spent just over $7.3 billion in the UAE. This includes our $1.5 billion equity investment in G42, more than $4.6 billion in capital expenses in advanced AI and cloud datacenters in the country, and more than $1.2 billion in local operating expenses and the cost of goods sold.   

• From the start of 2026 to the end of 2029, we will spend more than $7.9 billion in the UAE. This includes more than $5.5 billion in capital expenses for ongoing and planned expansion of our AI and cloud infrastructure, including new steps we will share publicly in Abu Dhabi this week. It also includes almost $2.4 billion in planned local operating expenses and the cost of goods sold.  

An investment in world-leading technology  

Some of our most important work involves exporting world-leading technology from the United States to the UAE. This includes the GPUs essential to power AI in our datacenters across the country that support the UAE’s people and institutions.  

Microsoft was one of the few companies during the previous administration to secure export licenses from the Commerce Department to ship GPUs to the UAE. In no small measure, this is because of the substantial work we did to meet the strong cybersecurity, national security, and other technology conditions required by these licenses. These licenses enabled us to accumulate in the country the equivalent of 21,500 Nvidia A100 GPUs, based on a combination of A100, H100, and H200 chips.  

Microsoft was also the first company this year under the Trump administration to secure export licenses from the Commerce Department to ship GPUs to the UAE. Approved in September, these were based on updated and stringent technology safeguards. These licenses enable us to ship the equivalent of 60,400 additional A100 chips, in this instance involving Nvidia’s even more advanced GB300 GPUs.  

While the chips are powerful and the numbers are large, more important is their positive impact across the UAE. We’re using these GPUs to provide access to advanced AI models from OpenAI, Anthropic, open-source providers, and Microsoft itself. We’re supporting AI-enabled applications, including our Copilot applications, from a wide variety of local and international providers. And we’re partnering with G42 to support public and private sector organizations across the UAE economy, as well as consumers across the country.  

The UAE’s ranking in the Microsoft AI Diffusion Report published last week shows the country leading the world in per capita AI usage. With 59.4 percent of the population using generative AI, the UAE is ahead of Singapore, which is in second place at 58.6 percent. Beyond these two, no other country tops the 50 percent mark. Microsoft’s infrastructure challenge in the UAE is not a risk of getting ahead of demand but keeping pace with it.  

Investing in talent  

Microsoft’s investment in the UAE is not just about technology, it’s also about people. By cultivating AI talent and skilling individuals to develop, deploy, and use AI in a way that reflects the region’s unique needs, Microsoft is helping to ensure that the UAE remains on the leading edge of AI diffusion.  

Today, Microsoft’s presence in the UAE reflects this commitment. Our growing team includes almost 1,000 full-time employees and related staff representing 40 nationalities. Nearly 100 of our employees are engineers, supported by an Emirati partner ecosystem that has grown almost threefold in just two years, now with 1,400 firms employing nearly 45,000 professionals across the country.  

This year, we established the Global Engineering Development Center in Abu Dhabi to attract world-class tech talent to the UAE. Our engineers not only develop new products and services for Microsoft, but support institutions and businesses across the region so they can use AI and cloud technologies to transform their own operations. As we look to the future, we aspire to grow our engineering teams further and add a new focus on domain-specific AI models and applications that will propel advances in key scientific and technological fields.  

We also opened a new center for the Microsoft AI for Good Lab in Abu Dhabi, staffed by PhD level research talent specializing in large-scale AI models, vision-language models, and post training techniques. Backed by Azure compute grants allocated to partner organizations, this team collaborates with nonprofits, start-ups, researchers, academic institutions, and local businesses to address humanitarian challenges across the Middle East and Africa using AI. Already, the Lab has partnered with researchers to train large language models for low-resource languages, including those spoken in Malawi, Tanzania, Kenya, Uganda, and the Democratic Republic of the Congo – helping ensure that AI serves communities that risk being left behind in the AI age.    

As the Microsoft AI Diffusion Report underscores, people need digital proficiency to fully participate in an AI-driven economy. Without the right skills, AI risks deepening inequality rather than broadening opportunity. That’s why skilling is a core pillar of our investment in the UAE.   

Last November, we committed to skill one million people in the UAE by the end of 2027–and we’re well on our way to meet – and exceed–our goal. Last month at GITEX, Microsoft partnered with UAE government entities to launch an initiative to upskill 120,000 government employees across the federal government, Abu Dhabi, Dubai, and Sharjah. We will also skill 175,000 students and 39,000 teachers through collaborations with GEMS, Abu Dhabi Department of Education (ADEK), and the Knowledge and Human Development Authority (KHDA). We will announce new steps in this area on Thursday in Dubai.  

Talent is the engine of AI leadership. Attracting, nurturing, and building AI talent and know-how is essential to the UAE turning its vision of becoming a global leader into a reality.   

Strengthening trust   

Ultimately, the use of AI depends on trust. People and institutions need to trust that AI will be developed and deployed with responsible safeguards. They need to have confidence that the cybersecurity of AI chips, models, and services will be protected. And they rightly expect AI to serve the public broadly, with cause for optimism that AI will reach and support the Global South.  

Given the role of export controls and other trade issues, the flow of advanced GPUs and AI models also requires trust between nations. This in turn requires clear rules and agreements, coupled with effective compliance systems. And as always, trust between nations depends on strong relationships between its people, grounded in mutual respect and appreciation for each other’s cultures.  

We’re focused on supporting all these needs.   

One important part of this is the Responsible AI Future Foundation, or RAIFF. G42, Microsoft, and Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) founded this new foundation in Abu Dhabi in February to promote responsible AI standards and best practices in the Middle East and across the Global South. This foundation is advancing research on the technical and ethical elements of responsible AI and is developing frameworks to ensure ethical development and deployment of AI systems, accounting for cultural diversity.   

A second element comes to life through the first annual Abu Dhabi Global AI Summit, which began Sunday. Hosted by G42, Microsoft, the Responsible AI Future Foundation, and Eurasia Group’s GZERO Media, this Summit brings government ministers, private sector executives, and AI leaders together to discuss what’s needed to drive AI diffusion across the Global South. We meet at a time when there is a growing risk that uneven AI diffusion may widen the economic and societal gaps that divide the world. It’s imperative that governments, businesses, and non-governmental organizations collaborate and take new steps to promote broader access to AI.  

Both these elements build on a third and deeper initiative that Microsoft and G42 have advanced during the past two years. In conjunction with our $1.5 billion investment, Microsoft and G42 created last April a first–of–its–kind binding framework between two private companies. Developed in close consultation with the U.S. and UAE governments, this Intergovernmental Assurance Agreement (IGAA) ensures that both our companies meet or exceed U.S. standards in critical areas such as cybersecurity and physical security, export controls and technology transfer, data protection and responsible AI, and Know Your Customer (KYC) best practices.   

As we drafted the IGAA, we consulted not only leaders from government ministries in our two countries, but with members and staff of both political parties in both houses of Congress in Washington, D.C. We listened to feedback and adapted the IGAA to address their suggestions. And we’ve built a strong compliance infrastructure to implement these requirements based on industry best practices and auditing standards.   

All these steps help bolster mutual confidence and trust between our two governments. But trust between nations also grows through relationships among people. That’s why we’re advancing a fourth and new element this week, traveling to Abu Dhabi with a Seattle delegation of public and private leaders. The group includes a former Governor and local leaders in economic development, higher education, medical research, the non-profit community, and sports. The goal is to deepen understanding, exchange ideas, and explore solutions that can advance both regions.   

Looking to the future  

Microsoft is committed to the future of the UAE and a strong relationship between our two nations. We believe in the UAE’s long-term economic vision and the role the UAE and the U.S. continue to play together to support peace, stability, and growth across the Middle East.  

As we do everywhere we do business, Microsoft is committed to a broad perspective and long-term approach. Our work in the UAE has underscored the importance of connecting technology investments to initiatives to attract and develop the talent needed for a vibrant and self-sustaining tech ecosystem. And work to advance trust, which may seem peripheral to some, is in fact of central importance. From stronger business practices to broader international ties, trust is a critical catalyst for technology success at a local and global level.  

Technology is our business and we’re as excited as anyone by its potential. But we know that ultimately there is only one test that matters. It’s how our technology empowers others to achieve more. Like every public company, our shareholders rightly expect us to deliver value to customers in ways that enable us to continue to grow. But we also judge ourselves by whether we are generating local opportunities and growth that go well beyond ourselves. Across the UAE, we’re committed to passing this test.  

The post Microsoft’s $15.2 billion USD investment in the UAE appeared first on Microsoft On the Issues.