Large language models can code like pros but also raise security risks. Enter BlueCodeAgent: a smart defense that learns from testing, applies principled rules and hunts down sneaky risks. The result is safer code, fewer false alarms and a 12.7% boost in catching trouble before it strikes.

Microsoft’s annual Ignite conference kicks off at 8 a.m. PT Tuesday in San Francisco. The event, held for developers, IT professionals and partners, will feature keynotes from Microsoft leaders and big news announcements in the era of AI. For those who can’t join in person, you can watch the live stream and keep up with the announcements on our Ignite 2025 website.

Around the world, the dangers of extreme weather are a daily reality. In 2024, extreme weather displaced or disrupted the lives of more than 800,000 people worldwide — a reminder that accurate, timely forecasts aren’t just about data; they’re about people.

Microsoft’s AI foundation model Aurora pushes past traditional forecasting to predict hurricanes, air quality, ocean waves and more with greater accuracy. Trained on vast amounts of diverse data, it’s tuned to model the Earth system, not just the atmosphere. Going forward, Aurora is entering its next phase through research partnerships and will remain open-source.

“Vibe coding” is making app creation possible for anyone who can describe an idea, not just professional developers. As conversational AI tools democratize tech and speed innovation, experts say it’s important to balance creativity with thoughtful structure. Learn more about how AI is helping everyone from business professionals to experienced developers solve problems, save time and be more creative.

Around 2.5 million Brazilians have taken advantage of the chance to learn how to use AI through a new Microsoft initiative, announced in September 2024, called ConectAI. Working in collaboration with the Brazilian government, ConectAI aims to help skill 5 million Brazilians to use AI by the end of 2027 through free online courses covering everything from AI fluency and digital literacy to more advanced technical skills.

Minecraft Education and Code.org’s Hour of AI turns coding into an epic quest. In “The First Night,” students use MakeCode and Python programming to teach a blocky Agent to recognize patterns, classify resources and coordinate mini-agent helpers – all while battling zombies. It’s coding meets adventure, with AI as your ultimate sidekick.

Microsoft is giving teens a voice in shaping AI’s future with the new AI Futures Youth Council. Alongside this, fresh research shows teens want transparency on how their data is used, worry about being excluded if they can’t verify their age and have mixed feelings about companies using biometric and behavioral data.

U.S. farms are vital to the nation’s food supply, but rising production costs, falling crop prices and declining farmable land are squeezing margins across operations of all size. Now, Land O’Lakes and Microsoft are teaming up to build AI tools like Oz, a digital assistant that will use Land O’Lakes data to help farmers optimize operations, boost yields and reduce risk.

Microsoft has flipped the switch on a new class of AI datacenter – one that doesn’t stand alone but joins a dedicated network of sites functioning as an AI superfactory to train bigger AI models faster.

Generative AI powers apps and tools that help millions of people get things done faster and smarter, but it doesn’t serve everyone equally. Communities that aren’t well represented online often get left behind because most AI models learn from internet data. That means performance drops in many languages and cultural contexts, as well as in places where less-developed infrastructure makes access harder.

Project Gecko, led by Microsoft Research, aims to close those gaps so everyone — from farmers in Kenya to rural housewives in India — can tap into affordable, adaptable AI. These tailorable systems deliver vital expertise in local languages, with culturally relevant content and multiple modes including text, voice and video. Learn more on the Microsoft Research Blog.

Nearly 70% of companies are using AI in some way these days, but the real difference lies in how they’re using it. In a new Microsoft-commissioned IDC study, researchers found that companies at the front of the pack are adopting advanced tools to transform operations and drive growth. Read on to see how these leaders use AI to boost productivity, innovate faster and set new industry standards.

For years, conservationists have hiked deep into the forest to change batteries and collect important data. Now, with help from Microsoft’s SPARROW project, researchers can use a solar powered AI computing system to monitor biodiversity in remote areas. Watch Juan M. Lavista Ferres, corporate vice president and chief data scientist of the AI for Good Lab at Microsoft, give a TED Talk on the project.

Microsoft’s Secure Future Initiative is reshaping cybersecurity. The latest report shows steady gains across engineering pillars, new tools to protect customers and internal systems and stronger defenses against evolving threats. The goal is clear: accelerate innovation, enhance resilience and lead the industry toward a safer digital future

We’re in the middle of an AI transformation. Much has been said about what technology could do, but what about the people who are literally building it? Five Microsoft researchers shared their stories with Forward Future, a daily AI newsletter, about the big problems they’re tackling and the trends shaping their work in the middle of an ever-changing AI era.

Microsoft is expanding its Sovereign Cloud capabilities to meet rising global demands for control, compliance and AI scale. New capabilities include EU-based AI processing, Microsoft 365 Copilot localization in 15 countries and enhanced Azure Local services — all backed by in-country experts and a growing sovereignty ecosystem.

Cloud adoption is booming — but so are security risks. In 2024, organizations faced an average of nine cloud security incidents, up nearly 90% from 2023. As threats evolve and clouds get more complex, it’s time for security leaders to rethink their playbook. The International Data Corporation’s (IDC) latest insights spotlight five key strategies for staying ahead.

Microsoft’s Southeast Asia 3 cloud region is coming to the Malaysian city of Johor Bahru. It’s designed to power AI transformation, meet rising demand and reinforce the country’s regional digital leadership. When it launches, it’ll deliver Microsoft’s most advanced cloud services to support complex workloads and evolving needs across the region.

About 200,000 service members leave the military each year, going on to college or to work in a wide variety of fields. Colleges and universities are helping some veterans enter the burgeoning field of cybersecurity, where their military skills transfer well.

What kind of AI does the world really want? That’s the question Mustafa Suleyman, CEO of Microsoft AI, hopes to answer with his newly announced MAI Superintelligence Team. It’s all part of a larger move by Microsoft AI working toward Humanist Superintelligence (HSI) —incredibly advanced AI capabilities that always work for, in service of, people and humanity.

He says HSI is a common-sense approach to the field, designed to get “all the goodness of science and invention without the ‘uncontrollable risks’ part.” Keep reading for more on how Suleyman envisions ensuring humanity remains at the top of the food chain.

Imagine a world where an AI agent working for you could go out into the world and make decisions on your behalf, communicating with, for example, a company’s AI agent to book you a flight or order your groceries. This is one of the scenarios explored in Magentic Marketplace, an open-source simulation environment for exploring the numerous possibilities of agentic markets and their wide-reaching societal implications. Read on about other possible futures with AI agents, plus the pros and cons of each one.

Today’s cyberthreats are not just targeting individual enterprises — they are undermining the very foundations of our society, like hospitals, power grids and communication networks, Freddy Dezeure, Microsoft’s deputy CISO for Europe, writes in a new post. And as cybercriminals evolve, the European Union adopted two powerful new legislations that take aim at stopping them. Read more about the new measures and how they will work to ensure the success and continuity of governments and businesses.

Microsoft is bringing Copilot closer to home. By the end of 2026, 15 countries will get in-country data processing for Microsoft 365 Copilot — starting with Australia, the U.K., India and Japan. It’s a big win for trust, speed and control, especially for governments and regulated industries.

Pantone just dropped a color oracle that’s powered by Microsoft AI. The new Palette Generator whips up trend-savvy palettes in seconds, blending decades of color psychology with conversational smarts. It’s live in Pantone Connect, free to try and ready to turn your vague “moody tech sunset” vibe into a fully forecasted color story. 

AI is rapidly becoming the backbone of our world, promising unprecedented productivity and innovation. But as organizations deploy AI agents to unlock new opportunities and drive growth, they also face a new breed of cybersecurity threats. In a new post, Charlie Bell, EVP of Microsoft Security, writes about the three principles that leaders should consider when adopting agentic AI.

No more pet names with exclamation points — wonky passwords are a thing of the past with passkeys. Available now for many users, passkeys sync across your Windows desktop devices via Microsoft Password Manager in Edge. Sign in without passwords, using built-in security like fingerprints or facial recognition. 

The speed of AI adoption is outpacing the radio, Internet and smartphone — 1.2 billion users in under three years. But the spread isn’t equal: usage in the Global North is double that of the Global South. Without intentional action, this divide could shape who benefits from AI for decades. Read Microsoft Research’s AI Diffusion Report.

Generative AI is transforming cybersecurity — for better and worse. Microsoft’s 2025 Digital Threats Report shows attackers doubling down on AI to craft deepfakes, automate adaptive malware and sharpen phishing. With 66% of orgs building custom AI tools and 80% worried about data leaks, Microsoft’s new guide breaks down five threats every security team should know.

Kenya just launched KAISA — a national AI skilling alliance backed by the Kenya Private Sector Alliance (KEPSA) and Microsoft. The goal is to unify fragmented efforts across sectors, boost innovation and expand access to AI training. With Africa eyeing a $1.5 trillion AI windfall by 2030, Kenya is positioning itself as a leader in ethical, inclusive adoption.

Boo! Microsoft Edge just made life harder for internet scammers this Halloween: Scareware Blocker is now enabled by default on most Windows and Mac devices. The AI-powered feature spots full-screen scams before traditional threat intelligence catches them — without slowing down your computer.

Microsoft is expanding its TechSpark initiative with new partnerships at TitletownTech and the New Jersey AI Hub, aiming to accelerate scientific discovery through high-performance AI and cloud computing. By combining startup agility, academic expertise and AI firepower, the company hopes to broaden access to innovation and ensure breakthroughs reach beyond research labs and into local communities.