Search AI Products and News

The European Union recently released a draft of guidelines for the behavior of General-Purpose AI (GPAI) models, aimed at providing companies with a framework to comply and avoid hefty fines. The document is set to be finalized in May and outlines risk management guidelines, offering a blueprint for compliance and avoidance of significant penalties. GPAI refers to AI that has been trained with a total computational power exceeding 10²⁵ FLOP. Companies expected to be bound by the EU guidelines include OpenAI, Google, Meta, Anthropic, and Mistral.

Meta has released a mid-term update showing a significant increase in the adoption rate of its Llama series models, especially after the launch of Llama 3.1. The downloads of the Llama models on the Hugging Face platform are approaching 350 million, a tenfold increase from last year, attracting well-known companies such as Zoom, Spotify, AT&T, and Goldman Sachs. The development of open-source AI not only approaches the performance of closed models but also achieves widespread application at the enterprise level.

Meta's open-source model Llama 3 has cooled down in the market, which has intensified competition with closed-source models. Amazon AWS has reacted calmly, and enterprise clients are more inclined to favor Anthropic's closed-source model Claude. Microsoft also tends to recommend that engineering and data science teams use other models. This phenomenon has prompted Meta to consider forming a dedicated AI sales team to better serve enterprises. Open-source models face challenges in commercialization, with actual performance and commercial returns falling short of expectations.

OpenAI welcomes former Meta executive Irina Kofman, who will oversee the company's strategic plans with a focus on AI safety and preparedness, reporting to technology head Mira Murati. Kofman's extensive experience in AI product management will bring new perspectives and strategies to OpenAI. The company is actively recruiting executives from large tech firms to enhance its competitiveness, recently appointing former Instagram and Twitter executives as Chief Product Officer and Chief Financial Officer. Meanwhile, competition is intensifying.

Zuckerberg and Ek jointly criticized the EU AI regulations, arguing that these complex rules hinder the development of open source AI in Europe and limit innovation and developer vitality. They called for simplification and unification of regulations, emphasizing that open source AI creates a fair competitive environment. Europe should leverage its single diverse market advantage to avoid missing out on technological development opportunities. Although the EU legislation exempts open source AI, the conditions limit its potential, and the transparency requirements for datasets have little impact on open source AI. Zuckerberg also stated that he is willing to release a value of 10 if it is safe and beneficial to the public.

Snap and Meta are set to unveil their latest AR glasses prototypes next month, with Snap CEO Evan Spiegel releasing the fifth-generation Spectacles on September 17, and Meta's Mark Zuckerberg showcasing the AR glasses codenamed Orion on September 25. However, due to the current immaturity of AR technology, neither pair of glasses will be available for public sale. Snap will distribute a small production run through partners and developers.

Meta's security report reveals Russia's use of generative artificial intelligence (AI) to enhance online deception activities. While Russia has attempted to leverage AI to improve the effectiveness of its cyber interventions, the results have been minimal, mostly boosting the productivity and content generation capabilities of malicious actors. The report indicates that since Russia's invasion of Ukraine in 2022, these activities have primarily focused on undermining the credibility of Ukraine and its supporter countries. Meta predicts that as the US elections approach, Russia may target candidates supporting Ukraine.

Meta (formerly Facebook) and Universal Music Group (UMG) have expanded their long-standing music licensing agreement aimed at enhancing copyright protection while addressing solutions for the application of AI technology in the music field. The new agreement allows users to freely share songs from UMG's music library across Meta's platforms, including Facebook, Instagram, Horizon, Threads, and WhatsApp, ensuring that users can enjoy music while avoiding copyright infringement.

Meta has signed a multi-year partnership with Universal Music Group (UMG) aimed at expanding the use of music content across Meta's social applications, including Threads and WhatsApp, and addressing the impact of unauthorized AI-generated content on artists. This collaboration deepens the partnership that began in 2017, allowing users to create and share music content more widely on these platforms. The agreement seeks to address copyright issues, provide protection for artists, and enable music to circulate again on the platforms, fostering user engagement with audio content.

The VFusion3D AI model, developed in collaboration with the University of Oxford, can transform a single 2D image or text into high-quality 3D objects, having a profound impact on the fields of virtual reality, gaming, and digital design. This model effectively addresses the scarcity of 3D training data by generating synthesized 3D data through pre-trained video AI. In practical tests, VFusion3D performed excellently compared to leading systems, with over 90% of human evaluators preferring its generated 3D reconstructions. VFusion3D can transform input in just seconds.

According to multiple media reports, social media giant Meta is actively in talks to incorporate voices of several well-known actors and influencers for its AI products. This new feature may debut as early as this year's Meta Connect conference in September.

As the developer of the Llama foundation language model, Meta points out that the computing power required for future model training will significantly increase, with the expected training amount for Llama 4 being 10 times that of Llama 3. In light of this trend, Meta emphasizes the importance of building training capabilities in advance to avoid falling behind in the competition. Meta has released Llama 3, which has 80 billion parameters, and its upgraded version Llama 3.1, which reaches 405 billion parameters, making it its largest open-source model. To meet future model demands, Meta is

Meta's AI assistant falsely reported a nonexistent assassination attempt on former President Trump, leading to widespread discussion and regret from Meta executives. The misleading response was caused by programming limitations initially set to avoid addressing questions about Trump's assassination, but after these restrictions were removed, the AI still provided incorrect information in some instances. This phenomenon, known as 'hallucination,' is a common challenge faced by generative artificial intelligence. Similar issues have also plagued Google, whose search autocomplete feature was criticized for censoring Trump-related queries.

Meta has launched a new tool called 'AI Studio' on Instagram, allowing users in the U.S. to create personal AI avatars aimed at enhancing interaction with their fans. These AI avatars can communicate directly, reply to comments, and act as virtual representations of the user, accessible via web or app. Users can customize content, set interaction rules, and even enable automatic replies. The creation of AI avatars allows for customization, enabling them to resemble the user or generate entirely new characters for use across Meta's applications. This move aims to compete with Character.AI.

Meta has announced the shutdown of its celebrity chatbot project, which allowed users to interact with virtual avatars of celebrities. Despite significant promotion, the project is no longer in service. Meta's focus has shifted to AI Studio, a new tool that allows users to create their own chatbots. Initial projects included virtual characters such as Coco, Victor, and Amber, but they failed to meet expectations, particularly in attracting more celebrities, resulting in low returns on investment. Nevertheless, Meta remains committed to AI chat technology.

The U.S. Department of Commerce report supports open-weight generative AI models, such as Meta's Llama3.1, believing this will help small companies, researchers, non-profits, and individual developers access AI resources, fostering innovation and competition. This view aligns with that of Federal Trade Commission Chair Lina Khan, who emphasized that open models can promote more small businesses in the creative market. Assistant Secretary Alan Davidson stated that the openness of large AI systems significantly impacts competition, innovation, and risks, and called for monitoring of large AI models broadly.

Meta recently announced the launch of a new tool called AI Studio, allowing users in the U.S. to create their own AI versions on Instagram or online. The company stated that creators and business owners can use these AI profiles to interact with fans, conversing directly in chat threads and responding to comments.

At the 51st SIGGRAPH conference, NVIDIA CEO Jensen Huang and Meta CEO Mark Zuckerberg engaged in a fireside chat discussing the future of artificial intelligence and simulation. During their conversation, Huang praised Meta's open source large model strategy and emphasized the importance of open source for AI development. They showcased their friendship and shared vision for advancing the industry through a creative 'coat swap' ceremony.

The Prompt-Guard-86M model released by Meta is designed to defend against prompt injection attacks by restricting large language models from processing inappropriate inputs, thereby protecting system security. However, the model itself also exposes risks of being attacked. Research conducted by Aman Priyanshu found that by adding simple character spacing such as spaces or removing punctuation in the input, the model disregards prior security instructions, achieving an almost 100% success rate for attacks. This finding highlights the importance of AI security, despite Prompt

At this week's SIGGRAPH conference, Meta CEO Mark Zuckerberg showcased Segment Anything 2 (SA2) for the first time, an upgraded version of the groundbreaking image segmentation model the company introduced last year. The new model extends AI-driven segmentation technology to the video domain, demonstrating the remarkable progress made over the past year.