Norton has announced the launch of advanced deepfake protection capabilities, developed in collaboration with Intel for the latest generation of Intel Core Ultra processors.
The new feature enables Norton 360 customers using the most recent Intel-powered AI PCs to benefit from real-time, on-device detection of AI-generated video scams. The solution is integrated into Norton Genie Scam Protection and is currently available to users in the US, UK, Australia, and New Zealand.
Real-time detection
The integration is designed to offer immediate analysis and warnings for suspicious video material, aiming to prevent both personal and financial harm caused by increasingly sophisticated deepfake scams. Unlike previous cloud-based systems, this feature operates directly on the device, utilising the neural processing unit (NPU) within Intel’s Core Ultra processors.
According to Norton, on-device processing increases both the speed and privacy of scam detection efforts, as sensitive data does not leave the user’s device.
“From the early days of fighting email and web scams to now tackling sophisticated video deception, Norton has always stayed a step ahead continuously evolving its protection to outsmart new threats,” said Leena Elias, Chief Product Officer at Gen. “Now, with scam detection for videos powered by the latest Intel processors, we’re raising the bar again. Norton Deepfake Protection delivers real-time, private, and ultra-fast protection right on people’s devices, so they can feel confident they’re a step ahead of scammers and their latest AI-generated tricks.”
Audio and video analysis
The technology is described as taking a comprehensive approach to detecting scam content within videos. Rather than simply searching for technical anomalies, Norton Deepfake Protection is designed to analyse both the visible content and the underlying message. The AI model has been trained on a broad and expanding database of known scams, including fabricated financial offers, false giveaways, and cryptocurrency cons.
In addition to video analysis, the system also examines audio tracks for indicators of synthetic manipulation. Norton likens this approach to a team of expert sound analysts meticulously investigating recordings for subtle signs of tampering, treating each audio element as “a piece of evidence.”
Privacy and efficiency
Local device processing is central to both the privacy and efficiency of the new feature. By utilising the NPU in Intel’s hardware, scam detection results are generated rapidly and without transmitting any user data to the cloud.
“Today’s cyber threats demand real-time detection, which is why Intel is excited to partner with Norton to bring AI-powered cybersecurity directly to users’ devices.” said Carla Rodríguez, Vice President and General Manager, Client SW Enabling at Intel. “Together, we are demonstrating the power of local compute right on the AI PC, enabling instant detection of AI-generated scams in video content without compromising privacy or requiring cloud connectivity. We’re proud to help Norton deliver the protection that today’s digital landscape requires.”
Platform support and availability
The feature currently supports automatic detection for YouTube videos viewed in any browser and for Facebook videos in Chrome, with support for other browsers and social media platforms expected in future updates. Users of Norton’s Genie Scam Protection and Scam Protection Pro in Norton 360 products can access the deepfake detection feature on Intel Core Ultra-powered PCs.
Norton confirms that support will be expanded to additional platforms and devices over time, and that similar scam detection capabilities were previously introduced in the Genie AI Assistant for Norton 360 on mobile devices in July. This approach allows individuals who do not have the latest AI hardware to benefit from video manipulation detection as well.
Norton states that its deepfake protection technologies are part of an ongoing effort to adapt to evolving scam types and to distribute scam prevention technology more widely and quickly.
