There is a vital global conversation underway about how to best protect young people online. Governments worldwide are introducing a range of proposals, from restrictions on personalized feeds and screen time to outright social media bans. While these proposals share a crucial goal that we support, they often overlook a major practical hurdle: how platforms can safely and accurately verify a teen’s age.

For any of these proposals to succeed, apps must know the age of their users. But proving age on the internet remains a complex, industrywide challenge. Many teens don’t have traditional government IDs, and requiring people to upload sensitive personal documents to every individual app they download creates significant privacy risks. Furthermore, smaller or emerging platforms often lack the robust security infrastructure required to safeguard this data, which can inadvertently expose millions of people to security breaches.

Protecting young people online should not come at the expense of privacy. That’s why parents and safety advocates overwhelmingly support a simpler, more secure approach: centralizing age verification and parental consent in the app store. By handling age verification once at the device level, we can provide young people with consistent, age-appropriate experiences across the many apps they use while keeping their personal data safe.

Unintended Consequences of Bans

We understand the immense pressure lawmakers face to act, and we respect every government’s right to decide what’s best for their citizens. But any truly viable safety proposal must solve the underlying challenge of how to uniformly and accurately understand a person’s age. 

Australia’s under-16 social media ban highlights just how complex this logistical piece remains. Because the policy was introduced without an established, privacy-preserving method for age verification, it has led to the unintended consequences safety experts feared: reports of teens bypassing inconsistent age checks, circumventing restrictions, and migrating to unmonitored apps and gaming sites that fall outside the scope of the ban.

This shift actually makes the internet less safe for young people. When teens access platforms through workarounds, they lose the built-in protections, like Teen Accounts on Instagram, Facebook, and Messenger, specifically designed to keep them safe. These concerns are shared globally; a joint statement from over 370 international academics and privacy experts warned that enforcing broad bans carries massive risks to privacy and autonomy if age verification isn’t built on a coherent, secure foundation. They’re right. If we get this wrong, we create entirely new risks for everyone online.

While Meta will always comply with local laws, current legislative proposals simply do not address these complex age issues — and that should give global lawmakers pause.

The Path Forward

There is a practical framework that directly answers the complex logistical challenges I’ve laid out: centralizing age verification and parental approval at the app store level. App stores are already the gateway through which teens access every app on their phones. And we don’t have to start from scratch. Apple and Google already collect age information when a parent sets up their teen’s phone, and they already have systems in place to obtain parental approval before teens can make purchases. We’re simply asking that this same mechanism be extended to all app downloads. By verifying a person’s age just once at this device level, the phone itself acts as a single, secure checkpoint. This allows parents to seamlessly approve or deny downloads across all platforms simultaneously, removing the need for people to upload sensitive personal documentation to dozens of individual apps.

Over the past year, more than half of U.S. states have introduced app-store age legislation of this kind, with Texas, Utah, Louisiana, Alabama, and California already enacting versions of it — driven largely by parents. Momentum is also building in Washington, where the App Store Accountability Act is advancing through Congress.

Public support is clear: polls show that 85% of American parents support requiring app stores to verify age and get parental approval before teens download apps. Eighty-two percent of Australian parents and nearly 75% of parents across eight European countries support parental approval. Major industry players like Match Group, X, Snap, and Pinterest have also endorsed this approach.

The Core Question

We will continue to comply with laws around the world as they evolve. But the conversation should always come back to the same core question: how will we, as a society, reliably verify age on the internet? Until we answer that question honestly, everything else is a workaround.

The post Why Social Media Bans Alone Can’t Solve the Age Verification Dilemma appeared first on Meta Newsroom.

Imagine you’re visiting a new city and want to find a restaurant that impresses your vegan in-laws. Using voice conversations on the Meta AI app, you ask, “Hey Meta, what are the best vegan options around?” 

Within seconds, Meta AI — powered by Muse Spark — responds with a list of local vegan restaurants, a short description of each restaurant’s vibe, and a map showing you exactly where the restaurants are. It’s a quick and seamless interaction that feels effortless, but behind that brief exchange are layers of calculations enabled by compute power.

What Is Compute Power?

Simply put, compute power is the measure of how much work a computer chip can do and how fast it can do it — like horsepower in a car engine. Compute power is measured in FLOPS: floating-point operations per second, or the number of calculations that a chip can perform in one second. FLOPS measure the speed of compute and gigawatts measure the scale of it, or how many chips you can keep running at once. 

When you ask Meta AI to find a vegan restaurant, it runs billions of calculations in just a few seconds. Your voice is captured, converted from sound waves into text, and routed to computers or servers inside a data center. From there, a large language model (LLM), and the result is delivered right to your ear. 

Even simple actions, searching for a local barbershop on Instagram, require layers of computation: understanding language, processing your query, scanning an index, generating results, and delivering it back to you, all before your thumb leaves the screen. All of this processing power is made possible by processing chips inside the servers inside our data centers.  

So why does the future of AI depend on compute? Here’s a closer look.

The Building Blocks of Compute

Compute is an abstract concept, but it’s delivered by physical chips. Different chips are designed to handle different types of calculations and workloads. 

• Central Processing Units (CPUs) are the processors in computers that make AI training and inference possible. Traditional CPUs were designed to handle tasks one at a time, and are great at managing network traffic, running application logic, and coordinating workflows across systems.

if (typeof jetpackSlideshowSettings === 'undefined') { window.jetpackSlideshowSettings = { spinner: 'https://about.fb.com/wp-content/mu-plugins/jetpack/modules/shortcodes/img/slideshow-loader.gif', speed: '4000', label_prev: 'Previous Slide', label_stop: 'Pause Slideshow', label_next: 'Next Slide' }; } jQuery(document).ready(function($) { $('.jetpack-slideshow-noscript').remove(); var container = $('#gallery-48593,48592,48589-9-slideshow'); if (!container.data('processed')) { var slideshow = new JetpackSlideshow( container, container.data('trans'), container.data('autostart') ); slideshow.images = container.data('gallery'); slideshow.init(); container.data('processed', true); } });

 

• Graphics Processing Units (GPUs) are processors that were initially designed for rendering graphics, but are also great at doing thousands of calculations simultaneously — the exact kind of processing we need to power AI. Training a model to understand languages, recognize images, or even engage in conversation requires large-scale calculations running simultaneously, repeatedly, and for weeks or even months on end. Both CPUs and GPUs exist in consumer products like laptops and smart phones, but the ones in data centers are built to be much more powerful.

• Custom chips are processors built for specific workloads, designed to maximize efficiency for tasks like ranking, recommendations, and generative AI. Meta has developed Meta Training and Inference Accelerator (MTIA), a family of custom silicon chips designed specifically for our AI workloads. Mainstream GPUs are typically built for large-scale AI training then applied less cost-effectively to other AI workloads like inference. MTIA takes a different approach: to prepare for the growth in AI inference demand, we build chips that are optimized for our inference workloads but are also able to support all workloads including training. This offers flexibility and efficiency that’s unmatched by any combination of general-purpose chips and enables us to innovate for the future of AI.

if (typeof jetpackSlideshowSettings === 'undefined') { window.jetpackSlideshowSettings = { spinner: 'https://about.fb.com/wp-content/mu-plugins/jetpack/modules/shortcodes/img/slideshow-loader.gif', speed: '4000', label_prev: 'Previous Slide', label_stop: 'Pause Slideshow', label_next: 'Next Slide' }; } jQuery(document).ready(function($) { $('.jetpack-slideshow-noscript').remove(); var container = $('#gallery-48590,48588,48595-10-slideshow'); if (!container.data('processed')) { var slideshow = new JetpackSlideshow( container, container.data('trans'), container.data('autostart') ); slideshow.images = container.data('gallery'); slideshow.init(); container.data('processed', true); } }); How Does Compute Power Meta’s AI?

At Meta, we’re building a global network of AI-optimized data centers, each designed with the flexibility to support both our AI workloads and the other workloads that are central to our apps and services. We believe that building at this scale requires a diversified approach to infrastructure. That’s why we’re sourcing silicon from a range of partners to ensure the right chips are matched with the right workload, allowing us to build and deliver new AI experiences at a faster pace. 

Our custom MTIA silicon is an essential part of our efforts. We’re developing and deploying four new generations of chips within the next two years to support ranking, recommendations, and generative AI workloads. In April, we announced an expanded partnership with Broadcom to co-develop multiple generations of MTIA chips.

if (typeof jetpackSlideshowSettings === 'undefined') { window.jetpackSlideshowSettings = { spinner: 'https://about.fb.com/wp-content/mu-plugins/jetpack/modules/shortcodes/img/slideshow-loader.gif', speed: '4000', label_prev: 'Previous Slide', label_stop: 'Pause Slideshow', label_next: 'Next Slide' }; } jQuery(document).ready(function($) { $('.jetpack-slideshow-noscript').remove(); var container = $('#gallery-48596,48594,48591-11-slideshow'); if (!container.data('processed')) { var slideshow = new JetpackSlideshow( container, container.data('trans'), container.data('autostart') ); slideshow.images = container.data('gallery'); slideshow.init(); container.data('processed', true); } });

And earlier this year, we announced a partnership with Arm to co-develop the Arm AGI CPU, — the first data center processor specifically designed to handle the massive amount of data movement demanded by AI workloads. We’ve also announced partnerships with industry leaders AWS, AMD, and NVIDIA to supply chips for our compute portfolio. 

These partnerships will enable us to continue innovating and building AI tools for the future. We recently announced Muse Spark, our most advanced AI model to date and the first LLM built by Meta Superintelligence Labs. Muse Spark is natively multimodal, processing voice, text, and images together. What makes it possible is compute at every level — from training models across thousands of GPUs to supporting billions of inferences each day on custom MTIA chips — and all of it running through efficient networks of servers at data centers around the world.  

The demand for more powerful and efficient compute will only accelerate. As AI continues to become more capable, personal, and integrated into people’s lives, we’ll keep building the infrastructure needed to power it.

The post Infrastructure Explained: Compute Power appeared first on Meta Newsroom.

Meta and Reliance Industries today announced a significant expansion of our strategic partnership with an agreement for an AI-enabled data center in India. Located in Jamnagar, Gujarat, this investment reaffirms Meta’s deep commitment to India, bringing infrastructure that powers our products and AI capabilities needed to deliver personal superintelligence to one of our largest and fastest-growing communities globally. As part of the agreement, Reliance will build a data center with 168 MW capacity, which Meta will lease, with options to scale.

“We’re proud to be working with Reliance to build our first AI-enabled data center in India. This world-class facility in Jamnagar will help us scale our AI infrastructure globally while deepening our long-term investment in India’s economy.”

– Mark Zuckerberg, Founder and CEO, Meta

Building in India

Meta is investing aggressively to expand our capacity footprint to support our technologies, services, and AI ambitions, which serve billions of people worldwide. India’s rapidly growing tech-forward digital economy, its massive user base, and the strength of our partnership with Reliance make India an ideal place to invest. 

Meta is leasing capacity at a new Reliance data center in Jamnagar, India, powered by renewable energy and cooled with desalinated seawater. Meta will cover the full cost of the energy and water supporting the facility. This investment is  a significant milestone in Meta’s global infrastructure expansion and deepens our long-standing strategic partnership with Reliance — one that spans connectivity, commerce, and AI innovation in one of the world’s most dynamic digital markets.

Jamnagar is a strategic location, and Reliance is developing one of the largest data center campuses in the world there, with access to the significant energy resources needed to power advanced AI-enabled infrastructure. The facility’s first phase will deliver 168 MW of capacity, with an option to scale. Paired with our extensive network investments, including Project Waterworth, the world’s longest subsea cable system, we will bring industry-leading connectivity to the region and serve India’s community with speed and quality.

A Partnership Built Over Time

Meta and Reliance Industries have built a strong partnership over the years, and this investment reflects a shared vision to accelerate India’s digital future. In 2020, we made a landmark $5.7 billion investment in Jio Platforms, accelerating connectivity and empowering small business growth across India. We deepened the collaboration through a joint venture bringing Meta’s open-source AI models to Indian enterprises and developers. Today’s data center agreement marks the next chapter — extending our partnership to the physical infrastructure powering Meta’s products and AI capabilities in one of the world’s largest and fastest-growing digital economies.

“This partnership with Meta marks a transformative moment for India’s digital infrastructure. Building India’s first built-to-suit AI data centre for a global technology leader of Meta’s scale demonstrates India’s readiness to be at the forefront of the global AI revolution. At Reliance, we are committed to building world-class digital infrastructure that will power the next generation of AI innovation — not just for India, but for the world.”

– Mukesh D. Ambani, Chairman and Managing Director, Reliance Industries Limited

Supporting Clean Energy Across India

Meta is also proud to announce we have contracted nearly 1 GW of new clean and renewable energy in India through agreements with two leading clean energy providers:

• CleanMax: 837 MW of new solar and wind projects in Rajasthan and Karnataka. This brings the cumulative capacity announced with CleanMax to over 900 MW.
• Fourth Partner Energy: 88 MW of new solar and wind projects across Tamil Nadu, Karnataka, Maharashtra, and Uttar Pradesh.

Coupled with the work by Reliance to support the Jamnagar data center with renewable energy, these additional agreements represent a meaningful commitment to India’s clean energy ecosystem. This helps ensure that Meta’s growing infrastructure in the country is supported by renewable power and will help address Meta’s value chain emissions in the region. This is consistent with our global goal of matching all operations with 100% clean and renewable energy.  

The post Meta Partners With Reliance on AI-Enabled Data Center in India appeared first on Meta Newsroom.

Meta personalizes the experiences of 3.5 billion people who use our apps and services every day, helping them accomplish what they come to our apps to do faster and more easily. Personalized ads help people discover content that matters to them — like a local ice cream shop or new hiking shoes — and keep our services free. Businesses often share information about people’s activity on their sites with us to make ads more relevant. For years, we’ve offered tools that help people understand how this works and manage their experience, and now we’re sharing some important updates to our approach.

Using Your Activity From Other Businesses for Better Personalization

We’re updating how we use information that other businesses already share with Meta. We already use this data — like games you play or purchases you make on other websites — to make the ads you see more relevant. In the future, we’ll use this information to personalize other parts of your experience, including the content you see in your Feed and AI responses. 

Your Activity off Meta Technologies Control Is Going Away

We currently offer two settings that let people manage activity shared with us from other businesses: “Your activity off Meta technologies” and “Activity from other businesses.” Instead of maintaining two settings that cover similar ground, we’re streamlining our controls and will no longer offer the “Your activity off Meta technologies” setting that lets you disconnect activity that businesses share with us from your account. As part of this update, we’re also expanding the “Activity from other businesses” setting, which lets you control how we use this data to personalize your experience.

You Decide How Your Activity Data Is Used

The setting for “Activity from other businesses” (formerly known as “Activity information from ad partners”) allows you to manage how we use this activity data to personalize the content we show you, both for ads and now non-ads content. If you allow us to use this data to show you personalized content, the ads and other content you see will be more relevant. If you don’t, we won’t use this information to show you more relevant ads or other content.

What’s Not Changing

We aren’t collecting any new data as part of this update. This is about using information that businesses already send to us to further improve your experience. For example, if you’ve recently purchased a tent online, you might see more Reels about camping. 

These control and data-use changes will go into effect in the US and a number of other countries next month with more countries to follow. For the latest information, please visit our Help Center.

The post Better Personalization and Changes to Controls for Your Activity From Other Businesses appeared first on Meta Newsroom.

Meta is proud to be partnering with the National Urban League, the Associated Builders and Contractors, and CBRE, as well as community partners across the country including, the US Hispanic Chamber of Commerce, STRIVE, Boone County Economic Development Corporation (IN), Richland Parish Chamber of Commerce (LA), Workforce Solutions Borderplex (TX), and Ohio Chamber of Commerce (OH).

We are launching America’s Workforce Academy (AWA) to build upon the huge demand we saw for Meta’s first major initiative of this kind, Level-Up, Meta’s fiber installation training program that received 35,000 applications in the first seven days.

The United States labor market needs hundreds of thousands of fiber technicians, welders, plumbers, electricians and other skilled trade workers. At Meta, we see this as an incredible opportunity for these American heroes to power America’s future. 

if (typeof jetpackSlideshowSettings === 'undefined') { window.jetpackSlideshowSettings = { spinner: 'https://about.fb.com/wp-content/mu-plugins/jetpack/modules/shortcodes/img/slideshow-loader.gif', speed: '4000', label_prev: 'Previous Slide', label_stop: 'Pause Slideshow', label_next: 'Next Slide' }; } jQuery(document).ready(function($) { $('.jetpack-slideshow-noscript').remove(); var container = $('#gallery-48727,48728-15-slideshow'); if (!container.data('processed')) { var slideshow = new JetpackSlideshow( container, container.data('trans'), container.data('autostart') ); slideshow.images = container.data('gallery'); slideshow.init(); container.data('processed', true); } });

 

If you want to help build America’s future, you can apply to America’s Workforce Academy here.

Dina Powell McCormick, Meta President and Vice-Chairman, said: “The AI revolution is bringing change but also historic opportunities. Skilled workers electrified rural America one pole at a time. They manned the factories that built the arsenal that won World War II. Now a new generation will pour the foundations and lay the fiber that secures American strength in this new age.”

Mike Rowe, CEO of the mikeroweWORKS Foundation, said: “Closing America’s skills gap requires us to not only make a more persuasive case for the skilled trades in general, it requires us to completely rethink the way we train the next generation of skilled workers. America’s Workforce Academy does both. Workers are actually paid to learn. There is zero cost to them, no college debt and a fast certification, with a guaranteed job on the other end. This is an important step in the right direction, and one that I hope other companies will be inspired to take.” 

Marc H. Morial, National Urban League President and CEO, said: “America’s Workforce Academy represents the kind of bold, inclusive investment our economy urgently needs. At a time when too many Americans are searching for pathways to stable, family-supporting careers, this initiative opens doors, particularly for communities who historically have been excluded from opportunity. By removing cost barriers, providing industry-recognized credentials, and guaranteeing employment, AWA is helping to build a more equitable and resilient workforce for the future.”

Rachel Peterson, Vice President, Data Centers, Meta, said: “The AI infrastructure we’re building today requires an incredible workforce to make it a reality. America’s Workforce Academy is our commitment to building that workforce with the same ambition and long-term thinking we bring to the technology itself. America needs hundreds of thousands of skilled tradespeople — electricians, mechanics, fiber technicians and more — and this program creates clear, accessible pathways into those careers.” 

Michael Bellaman, ABC president and Chief Executive Officer, said: “America’s Workforce Academy is a transformational endeavor creating incredible opportunities, and ABC is proud to partner with Meta and CBRE to welcome all who want to build their career dreams in construction. This innovative talent pipeline solution addresses the industry’s ongoing workforce shortage by utilizing ABC’s existing, proven, nationwide education ecosystem. The sustained demand for data center construction technicians means the industry needs an all-of-the-above approach to grow the construction talent pool.”

Bob Sulentic, CBRE Chair and Chief Executive Officer, said: “We are excited to serve as a key partner for Meta on the management of America’s Workforce Academy — from candidate intake and qualification to hands-on training, in collaboration with the Associated Builders and Contractors. In this role, we are leveraging the full scope and expertise of CBRE to recruit, train, and deploy thousands of skilled workers who will support Meta in building out their AI infrastructure.”

The post America’s Workforce Academy: The Future Is for Everyone appeared first on Meta Newsroom.

Last year, WhatsApp made history by securing a landmark verdict and permanent injunction barring NSO Group — a spyware firm blacklisted for actions contrary to US national security — from targeting WhatsApp and its users ever again. The court was unequivocal: NSO violated federal and state laws against hacking. Today, we’re asking the court to hold them in contempt of that order.

Catching and Disrupting NSO’s Targeting Attempts

We successfully disrupted NSO-linked social engineering attempts, after investigating user reports. They tried to trick people into clicking on malicious links to drive them to external websites outside of WhatsApp, similar to previously reported 1-click phishing campaigns linked to NSO. We also caught them creating test accounts and groups on WhatsApp, which we took down. 

We are sharing threat indicators so that anyone can check if they were targeted by NSO-linked social engineering attempts across any platform — be it text message, email, WhatsApp message, or something else.

Spyware Is a National Security Threat

Since 2019, our case has shown that NSO continues to build spyware tools to target people’s devices. Its CEO confirmed in court that the company  looks for “vectors, or ways to access the phone” beyond WhatsApp, targeting browsers, operating systems, and other applications. 

No technology is off-limits to surveillance-for-hire firms, whose reported targets range from journalists to government officials, military personnel, and humanitarian organizations. 

When a malicious company on the US government’s Entity List continues to defy US courts, existing restrictions must remain firmly in place. Easing them would undermine US national security and put American companies and billions of people worldwide who depend on secure communications at risk.

No Company Can Fight Spyware Alone

When we originally discovered NSO’s 2019 attack, the Citizen Lab helped us provide additional notification to the people who were targeted. When the case first went to trial, our industry peers and other organizations supported it, strengthening the legal record. 

Last month, we were joined by 12 prominent civil rights organizations — a coalition of security researchers, privacy advocates, and digital rights experts — who filed their amicus briefs to fight NSO’s appeal against the permanent injunction. 

Today, we are beginning to deliver on our promise to support digital rights organizations working to defend people against spyware attacks by making a significant contribution to the Spyware Accountability Initiative (SAI). SAI supports dozens of organizations worldwide focused on forensic research, user support, and advocacy. 

For example, a Citizen Lab zero-day discovery led to an Apple security update for over a billion devices. This year, a Greek court issued the first-ever criminal conviction of spyware company executives, a case built on forensic evidence and investigative reporting by civil society.

This work is demanding, often dangerous, and consistently under-resourced compared to the spyware industry that continues to develop new exploits. We’re committed to doing our part to support this critical effort.

As always, WhatsApp users’ personal messages and calls remain protected with default end-to-end encryption. We encourage people to keep their apps and devices up to date and report suspicious activity so we can investigate and take action. For those who believe they may be targeted by sophisticated cyber attacks, we strongly recommend enabling strict account settings to harden their WhatsApp accounts even more.

Threat Indicators Malicious domains:

• hxxps://ikhwancast[.]com
• hxxps://ghazacast[.]com
• hxxps://fr24cast[.]com

The post Fighting Spyware: An Update From WhatsApp appeared first on Meta Newsroom.