Discover the Latest AI News📰

The AI search engine Perplexity faces a class-action lawsuit over claims its 'incognito mode' fails to protect user data. An anonymous plaintiff alleges the platform secretly shares conversations with Google and Meta, exposing sensitive financial discussions to advertisers. Investigations reveal the use of tracking technologies that capture detailed interaction history, raising serious concerns about digital privacy in AI tools.

Midea Group has quietly transformed from an appliance maker into an AI powerhouse, with over 13,000 intelligent agents now running daily operations. These digital workers handle everything from factory floors to customer service, creating what industry experts call 'the most advanced AI implementation in manufacturing today.' The company's Jingzhou washing machine plant demonstrates how AI coordinates complex production lines with minimal human oversight, while smart home products now anticipate consumer needs before they arise.

Xiaomi has introduced four subscription tiers for its MiMo large language model, starting at just 39 yuan per month. The plans offer developers transparent pricing across text, image, and audio capabilities. With its MiMo-V2-Pro ranking among global AI leaders, Xiaomi aims to simplify AI adoption while integrating with its smart ecosystem.

Didi Chuxing's AI travel assistant 'Xiao Di' has exploded in popularity, with active users growing 37 times since January. Nearly half of these users belong to the post-00s generation. As the Qingming holiday approaches, Didi anticipates transportation demand will spike dramatically, with inter-city trips expected to jump 40%. The company has upgraded its AI capabilities to handle complex holiday travel scenarios, shifting from simply finding rides to optimizing the entire journey experience through intelligent matching and real-time traffic analysis.

Xiaomi has introduced its first paid subscription plans for the MiMo large language model, marking a shift from free beta testing to commercial operation. With four pricing tiers starting at just 39 yuan per month, the plans offer access to Xiaomi's professional-grade AI models for both casual users and serious developers. The move reflects the tech industry's broader challenge of balancing AI accessibility with sustainable business models.

AI startup Ex-Human has taken legal action against Apple, alleging the tech giant unfairly removed its apps Botify AI and Photify AI from the App Store while withholding $500,000 in revenue. The company claims Apple provided vague reasons for the takedown, suggesting it might be clearing competition for its own Image Playground tool. This case highlights growing tensions between platform owners and AI developers over content moderation and revenue sharing.

Aliyun's Tongyi Lab has unveiled Wan2.7-Video, a groundbreaking AI-powered video creation tool that promises to revolutionize content production. Unlike traditional editing software, this innovative platform allows users to manipulate videos with unprecedented ease - from changing seasons with a click to altering character dialogue without reshoots. The tool addresses common creator frustrations by offering document-like editing simplicity combined with professional-grade results.

Google has introduced a more flexible pricing model for its Gemini API, offering five distinct service tiers tailored to different use cases. The new options range from budget-friendly batch processing to premium real-time response services, with some tiers offering up to 50% discounts. Developers can now choose between cost-saving flexible plans during off-peak hours or pay extra for guaranteed millisecond-level responses when speed is critical.

Ant Group's GPASS technology is transforming smart glasses into personal assistants. The latest update allows users to unlock shared bikes and pay for parking with simple voice commands - no phone needed. Imagine just saying 'Hello Qwen, I want to ride a bike' and you're rolling. The system combines voice recognition with Alipay's payment ecosystem, making urban mobility smoother than ever.

Google DeepMind has unveiled Gemma 4, its latest open-source AI model series, now available under the permissive Apache 2.0 license. The lineup includes four specialized models ranging from a flagship 31B parameter version to compact edge-compatible options. Performance metrics show dramatic improvements in math, coding and reasoning capabilities compared to its predecessor. Notably, the models feature built-in 'thinking mode' and native agent support, marking Google's strategic push to reclaim influence in the open-source AI space.

Artificial intelligence company Ex-Human has taken Apple to court, alleging unfair removal of its apps from the App Store without proper justification. The lawsuit claims Apple withheld about $500,000 in revenue while failing to provide specific evidence for the ban. The case highlights growing tensions between app developers and Apple's opaque review process, raising questions about potential anti-competitive behavior as Apple expands its own AI offerings.

The short drama 'The Peach Hairpin' has been removed from Hongguo Short Drama platform after creators used bloggers' faces without consent. The production company faces a 15-day suspension. This case highlights growing concerns about AI's role in content creation and the fine line between innovation and infringement in digital media.

Guangqi Honda's P7 electric SUV just received a game-changing software update that brings emotional AI to the cabin. The highlight? A voice cloning feature that lets your car speak in your family members' voices. Alongside this personal touch, the update adds smarter trip planning, natural conversations, and enhanced driving sounds - proving traditional automakers can still innovate in the EV space.

The Cursor team has unveiled version 3 of their innovative IDE, shifting software development into higher gear with autonomous AI agents. This major update introduces a unified workspace where multiple agents can collaborate across local and cloud environments. Developers gain new flexibility to switch contexts seamlessly while enjoying improved code review tools. The release signals a future where AI handles more coding tasks independently, though human oversight remains crucial.

Beijing has officially approved 15 new generative AI services under its registration system, marking a significant step in China's AI governance framework. These services, now legally compliant, must clearly display identification marks on AI-generated content to ensure transparency. The move balances innovation with regulation, offering citizens more tech choices while maintaining oversight.

A new children's safety alliance pushing for AI regulations has come under scrutiny after revelations that OpenAI was secretly bankrolling the effort. Multiple organizations joined what they thought was an independent coalition, only to withdraw upon learning of OpenAI's central role. Critics argue the tech giant's behind-the-scenes involvement undermines transparency in policymaking around AI and children.

Sarvam AI, an emerging star in India's tech scene, is closing a massive $350 million funding round with support from industry giants Amazon and NVIDIA. The investment could push the startup's valuation past $1.5 billion as it develops voice-focused AI models tailored for India's diverse languages. This deal highlights growing global interest in localized AI solutions beyond Western markets.

Meituan has unveiled LongCat-Next, a groundbreaking AI model that processes images, speech, and text as naturally as humans do. Unlike traditional systems that treat these separately, it converts everything into a common 'language' the AI understands natively. Early tests show it outperforms specialized models in reading documents, solving visual math problems, and maintaining top-tier language skills - all while being remarkably compact. The open-source release could change how we build AI that interacts with our physical world.

Voice AI leader ElevenLabs has stepped into the music creation arena with its new iOS app ElevenMusic. The platform lets users generate custom tracks using simple text prompts while offering social features like shared playlists and remix tools. Priced at $9.99 monthly for power users, this move signals ElevenLabs' push beyond voice technology into broader creative media as it competes with established players like Suno and Udio.

Google has unveiled Gemma4, its latest open-source AI model series featuring four variants with groundbreaking capabilities. The lineup includes efficient E2B and E4B models for edge devices and powerful 26B MoE and 31B dense versions that rank among the world's top open-source models. What makes Gemma4 special? It supports images, videos, and even real-time voice processing while being remarkably accessible for local deployment.