Hasta La Gadget!

Meta has admitted to CNBC that Instagram is experiencing an error that’s flooding users’ accounts with Reels videos that aren’t typically surfaced by its algorithms. “We are fixing an error that caused some users to see content in their Instagram Reels feed that should not have been recommended,” the company told news organization. “We apologize for the mistake.” Users have taken to social media platforms to ask other people whether they’ve also recently been flooded with Reels that contain violent and sexual themes. One user on Reddit said that their Reels pages was inundated with school shootings and murder. 

Others said they’re getting back-to-back gore videos, such as stabbings, beheadings and castration, nudity, uncensored porn and straight-up rape. Some said they still see similar videos even if they had enabled their Sensitive Content Control. Social media algorithms are designed to show you videos and other content similar to ones you usually watch, read, like or interact with. In this case, though, Instagram has been showing graphic videos even to those who haven’t been interacting with similar Reels, and sometimes even after the user has taken the time to click “Not Interested” on a Reel with violent or sexual content. 

The Meta spokesperson didn’t tell CNBC what exactly the error was, but some of the videos people have reported seeing shouldn’t have been on Instagram in the first place, based on the company’s own policies. “To protect users… we remove the most graphic content and add warning labels to other graphic content so that people are aware it may be sensitive or disturbing before they click through,” the company’s policy reads. Meta’s rules also state that it removes “real photographs and videos of nudity and sexual activity.”

This article originally appeared on Engadget at https://www.engadget.com/apps/meta-confirms-instagram-issue-thats-flooding-users-with-violent-and-sexual-reels-051631670.html?src=rss

Ahead of MWC 2025, HONOR has announced that its AI Deepfake Detection technology, first introduced at IFA 2024, will be available worldwide starting in April 2025. This technology is designed to help users identify manipulated media by providing real-time AI analysis of video and image content, offering immediate warnings against potential deepfake material.

Deepfakes are AI-generated synthetic media that alter images, videos, or audio to convincingly mimic real people. Using deep learning techniques such as Generative Adversarial Networks (GANs), these creations have become increasingly sophisticated, raising concerns about their impact on misinformation, fraud, and digital security. Notable cases include viral deepfake videos of Tom Cruise (on TikTok, 2021) and altered scenes from Back to the Future.

The rapid growth of deepfakes has prompted heightened concerns. According to the Entrust Cybersecurity Institute, a deepfake attack occurred every five minutes in 2024. A separate Deloitte 2024 Connected Consumer Study found that 59% of respondents struggled to distinguish between human-made and AI-generated content. Additionally, 84% of generative AI users advocate for content labeling to improve transparency.

Organizations like the Content Provenance and Authenticity (C2PA) are working on industry-wide standards to verify digital content authenticity. HONOR’s technology contributes to this effort by detecting subtle inconsistencies, such as:

• Pixel-level synthetic imperfections
• Border compositing artifacts
• Inter-frame continuity issues
• Anomalies in face-to-ear ratio, hairstyle, and facial features

Upon detecting manipulated content, users receive an immediate warning, helping them make informed decisions. Marco Kamiya of UNIDO, the United Nations Industrial Development Organization, has highlighted the importance of deepfake detection as a security measure for mobile devices and its role in combating digital fraud.

“In this current wave of digitisation, privacy security has become a top concern & priority in our everyday lives. As we intertwine more deeply with the digital world, a vast amount of our private information – from our daily whereabouts and financial information to social relationships and our personal preferences is stored in digital space.” said Kamiya.

HONOR is expected to share more information about the global deployment of its AI-driven technologies during a keynote at MWC 2025 where the company will unveil its HONOR ALPHA PLAN.

The event is scheduled for Sunday, March 2nd, at 16:30 CET. To stay updated on HONOR’s latest announcements at MWC, follow the official HONOR @ MWC 2025 page and social media channels.

HONOR Keynote Event @ MWC 2025

• Date: Sunday, March 2, 2025
• Time: 16:30-17:30 CET
• Venue: Hyatt Regency Barcelona Tower Hotel
• Location: Avinguda de la Granvia de l’Hospitalet, 144, 08907 Barcelona, Spain

HONOR To Release AI Deepfake Detection Globally In April 2025

, original content from Ubergizmo. Read our Copyrights and terms of use.

At MWC 2025 in Barcelona, HONOR is set to introduce its latest AI Agent solution under the umbrella of its  AI-driven HONOR ALPHA PLAN. This announcement marks a significant step in the company’s vision for integrating artificial intelligence into everyday technology.

The rise of AI agents promises to transform the way we access information and enhance our creativity. These intelligent systems are designed to support autonomous decision-making, manage complex tasks, and facilitate seamless collaboration between various AI agents. From digital assistants to physical robots, AI agents are expected to revolutionize both work and personal life.

HONOR has been actively advancing its AI initiatives. At IFA 2024, the company introduced its on-device AI Agent for the open ecosystem, alongside its AI-powered deepfake detection technology. These innovations highlight HONOR’s commitment to building responsible and efficient AI solutions.

This year, HONOR’s AI Agent solution will be unveiled alongside a comprehensive roadmap for global implementation. This milestone underscores the company’s ambition to play a leading role in the evolution of mobile AI, shaping the future of intelligent, interconnected devices. 

HONOR will share more details on its AI ALPHA PLAN during a keynote at MWC 2025. The event is scheduled for Sunday, March 2nd, at 16:30 CET. To stay updated on HONOR’s latest announcements at MWC, follow the official HONOR @ MWC 2025 page and social media channels.

HONOR Keynote Event @ MWC 2025

• Date: Sunday, March 2, 2025
• Time: 16:30-17:30 CET
• Venue: Hyatt Regency Barcelona Tower Hotel
• Location: Avinguda de la Granvia de l’Hospitalet, 144, 08907 Barcelona, Spain

MWC 2025 attendees are welcome to visit HONOR’s booth at Stand 3H10 in Hall 3 at Fira Gran Via to get their hands on the latest innovations.

HONOR to Unveil AI Agent Solution at MWC 2025

, original content from Ubergizmo. Read our Copyrights and terms of use.

Meta has admitted to CNBC that Instagram is experiencing an error that’s flooding users’ accounts with Reels videos that aren’t typically surfaced by its algorithms. “We are fixing an error that caused some users to see content in their Instagram Reels feed that should not have been recommended,” the company told news organization. “We apologize for the mistake.” Users have taken to social media platforms to ask other people whether they’ve also recently been flooded with Reels that contain violent and sexual themes. One user on Reddit said that their Reels pages was inundated with school shootings and murder. 

Others said they’re getting back-to-back gore videos, such as stabbings, beheadings and castration, nudity, uncensored porn and straight-up rape. Some said they still see similar videos even if they had enabled their Sensitive Content Control. Social media algorithms are designed to show you videos and other content similar to ones you usually watch, read, like or interact with. In this case, though, Instagram has been showing graphic videos even to those who haven’t been interacting with similar Reels, and sometimes even after the user has taken the time to click “Not Interested” on a Reel with violent or sexual content. 

The Meta spokesperson didn’t tell CNBC what exactly the error was, but some of the videos people have reported seeing shouldn’t have been on Instagram in the first place, based on the company’s own policies. “To protect users… we remove the most graphic content and add warning labels to other graphic content so that people are aware it may be sensitive or disturbing before they click through,” the company’s policy reads. Meta’s rules also state that it removes “real photographs and videos of nudity and sexual activity.”

This article originally appeared on Engadget at https://www.engadget.com/apps/meta-confirms-instagram-issue-thats-flooding-users-with-violent-and-sexual-reels-051631670.html?src=rss

Meta has admitted to CNBC that Instagram is experiencing an error that’s flooding users’ accounts with Reels videos that aren’t typically surfaced by its algorithms. “We are fixing an error that caused some users to see content in their Instagram Reels feed that should not have been recommended,” the company told news organization. “We apologize for the mistake.” Users have taken to social media platforms to ask other people whether they’ve also recently been flooded with Reels that contain violent and sexual themes. One user on Reddit said that their Reels pages was inundated with school shootings and murder. 

Others said they’re getting back-to-back gore videos, such as stabbings, beheadings and castration, nudity, uncensored porn and straight-up rape. Some said they still see similar videos even if they had enabled their Sensitive Content Control. Social media algorithms are designed to show you videos and other content similar to ones you usually watch, read, like or interact with. In this case, though, Instagram has been showing graphic videos even to those who haven’t been interacting with similar Reels, and sometimes even after the user has taken the time to click “Not Interested” on a Reel with violent or sexual content. 

The Meta spokesperson didn’t tell CNBC what exactly the error was, but some of the videos people have reported seeing shouldn’t have been on Instagram in the first place, based on the company’s own policies. “To protect users… we remove the most graphic content and add warning labels to other graphic content so that people are aware it may be sensitive or disturbing before they click through,” the company’s policy reads. Meta’s rules also state that it removes “real photographs and videos of nudity and sexual activity.”

This article originally appeared on Engadget at https://www.engadget.com/apps/meta-confirms-instagram-issue-thats-flooding-users-with-violent-and-sexual-reels-051631670.html?src=rss