Strengthening protections against non-consensual intimate imagery

When intimate images are shared without consent-whether real or AI-generated-the harm is immediate, deeply personal, and often long-lasting. It can affect someone's sense of safety, dignity, and control, both online and offline. Protecting people from harms like non-consensual intimate imagery (NCII) has long been a priority for Microsoft. And as technology advances, our response continues to evolve to tackle very real challenges like the proliferation of highly realistic synthetic imagery. With the US Take It Down Act coming into force this month, establishing new federal protections against the spread of NCII, it's important to share how we're evolving our approach: making it easier to report harm, taking new steps to detect known NCII, and enabling more effective enforcement across our services.

Expanding protections across Microsoft services

Our goal is to make it simpler for individual s, or t heir representatives, to report violative content to Microsoft. We have strengthened our global reporting processes for NCII with a more i ntuitive form , with clear options to describe harm, including both real and AI - generated images. These changes are designed to ease the burden f or people in a distressi ng moment and enable faster, more effective action by our teams. Microsoft's NCII policy is applied consistently across real and synthetic content , recognizing that the harm to individuals is the same , regardless of how an image was created. To report content on Mic ro so ft es , hit " encounter the content .

We also want to proactively detect and prevent the spread of known NCII by w orking with a reporting platform that enables individuals to create a digital "fingerprint," or hash, of their images. Two years ago, provided StopNCII.org with a new version of PhotoDNA that enables victims to create a hash without an image ever leaving their device . This can then be used by StopNCII.org partners to detect and remove matching NCII content across platforms , allowing industry to work together to prevent re-sharing and protect individuals' privacy . We have been piloting the use of these hashes in Bing since September 2024 .

We have now expanded our use of validated StopNCII.org hashes across Microsoft consumer services, including Teams Free, OneDrive, and Xbox. We will implement these changes carefully to advance effectiveness and accuracy-accelerating removals, automating where appropriate, maintaining human review for reported cases, and providing clear, accessible paths for users to appeal decisions.

Enhancing our collective response to this harm

No single company can address NCII alone. It requires coordination across industry, governments, and civil society. Microsoft will continue working with partners to improve shared tools and approaches that help prevent this content from spreading. We will also continue to advocate for clear, effective policies that protect victims, support innovation, and strengthen accountability across the ecosystem.

We will also continue to advocate for policies that support efforts to advance laws that prevent and deter image-based abuse. Microsoft advocated in support of the US Take It Down Act and welcomes the European Union's work to strengthen protections against "nudification" apps, alongside global efforts to criminalize this misuse of technology. We are closely tracking Ofcom's recent announcement that new measures will be required under the UK Online Safety Act to address illegal NCII harms. We believe our proactive work in this area will help us maintain trust with survivors, users, and regulators, among others.

Speed, clarity, and trust matter for people affected by intimate image abuse. When someone reaches out for help, we will strive to respond quickly, respectfully, and effectively. Our goal, though, is to invest in technologies and partnerships that reduce the likelihood of harm. We have joined forces with Childnet, a UK NGO that aims to safeguard children online, and created educational materials to prevent the misuse of AI to create intimate imagery among teens. These materials have now been released in the UK, as well as localized with partners in Singapore, South Korea, and Japan.

I am proud to learn from our digital safety team, which is carefully charting our path, and from the many industry and community leaders contributing to this work. This is an evolving challenge. We are committed to the journey, grounded by the voices of experts and survivors.