Discover the Latest AI News📰

DeepMind Intelligence has unveiled PhysBrain 1.0, a groundbreaking AI model that gives robots human-like understanding of physical laws. Unlike previous systems that simply mimic actions, this technology enables machines to predict and adapt to real-world environments. Developed by Beijing Zhongguancun College researchers, it could revolutionize how robots operate in unpredictable settings.

A groundbreaking initiative is offering financial aid and career retraining to workers impacted by AI automation. The program provides $1,000 monthly payments for a year, along with guidance for transitioning to more secure fields like healthcare. With tech jobs shrinking due to tools like GitHub Copilot, this pilot could become a model for addressing AI's growing impact on employment.

OpenClaw founder Peter Steinberger reveals stark contrasts in AI adoption between China and the U.S. While Chinese companies mandate AI tools for employees, some American firms ban them over security concerns. Steinberger shares insights from his Bloomberg interview, predicting a future where personal AI agents handle complex tasks across devices.

Anthropic is introducing dynamic pricing for its Claude AI, mirroring utility companies' peak-hour pricing models. During busy periods from 5-11 AM Pacific Time, users will burn through their conversation quotas faster - though weekly limits remain unchanged. About 7% of power users will notice the change most. The move aims to smooth out demand spikes while giving developers incentives to schedule heavy tasks during off-peak hours.

Google has expanded its Live Translate feature to iOS devices, allowing users with any headphones to experience near-real-time translation powered by Gemini AI. The technology now handles idioms and slang more naturally across 70 languages, moving beyond robotic translations. This marks a strategic shift from hardware-dependent features to AI-driven accessibility.

AI company Cohere has thrown its hat into the speech recognition ring with Transcribe, a lightweight open-source model designed for smartphones and edge devices. Supporting 14 languages and outperforming competitors like Alibaba's Qwen3, this 2-billion parameter model could change how we interact with voice assistants while addressing privacy concerns. The move signals Cohere's strategic shift from text generation to building complete AI agent capabilities.

Taotian Group is making waves in the e-commerce sector by equipping its interns with premium AI tools typically reserved for full-time staff. The company's new 'AI Productivity Plan' provides free access to advanced platforms like Wukong and Qoder, along with generous computing quotas. This bold move signals Taotian's commitment to building an AI-native workforce from the ground up, with technical roles dominating their 2027 internship program.

OpenAI's ChatGPT advertising pilot has surpassed $100 million in annualized revenue within six weeks of launch, far exceeding expectations. The program, currently limited to the U.S., keeps ads completely separate from conversations while protecting user privacy. With expansion plans for Australia, Canada and New Zealand underway, this success signals AI's growing role in reshaping digital advertising.

Wikipedia has officially banned the use of large language models to create or rewrite articles, marking a significant shift in policy. The decision, supported by an overwhelming majority of volunteer editors, aims to protect the platform's reputation for accuracy and reliability. While AI can still assist with basic editing suggestions and translations under strict guidelines, introducing new AI-generated facts is strictly prohibited. The move comes as editors grow increasingly concerned about AI 'hallucinations' compromising Wikipedia's integrity.

Tencent has pulled back the curtain on its ambitious Agent product ecosystem, playfully dubbed 'Shrimp Farm'. This comprehensive platform combines Tencent's Hunyuan model with other leading AI systems, offering businesses and developers a versatile toolbox for creating intelligent assistants that go beyond simple chat. With deep integration across Tencent's product suite and robust security measures, the system aims to transform how we interact with AI in daily work scenarios.

OpenAI has launched a game-changing plugin service for its Codex platform, allowing developers to package and share skills and configurations with just one click. This innovation streamlines team collaboration by enabling instant synchronization of development setups. The move comes as Codex surpasses 1 million users, with usage doubling since the latest update. The new plugin marketplace already integrates popular tools like Slack and Figma, signaling OpenAI's push toward standardized AI-assisted programming.

OpenAI has abruptly halted plans for a controversial 'adult mode' in ChatGPT and shut down its Sora video generation model. The moves come as part of a broader strategic shift away from consumer-facing projects toward enterprise solutions. Industry analysts suggest the company is responding to competitive pressure from Anthropic's growing foothold in business applications.

In a bold move to retain its top hardware talent, Apple is doling out hefty bonuses—some reaching $400,000—to key iPhone designers. The tech giant's counteroffensive comes as OpenAI and other rivals aggressively poach Apple's engineering veterans. With over 40 former employees already defecting to OpenAI alone, this escalating talent war reveals how crucial hardware expertise has become in the AI arms race.

Wikipedia has taken a firm stance against AI-generated content, voting overwhelmingly to prohibit its use in articles. The new policy strengthens existing guidelines, banning not just AI-written articles but also AI-rewritten content. While editors can still use AI for suggestions, all final content must be human-verified. This move aims to protect Wikipedia's accuracy and human editorial control as AI tools become more prevalent in content creation.

AI powerhouse Anthropic is gearing up for what could be one of tech's biggest IPOs this October, potentially beating rival OpenAI to the public markets. With Wall Street giants lining up to underwrite and tech titans like Google and Amazon backing it, the company just cleared a major legal hurdle that had threatened its billion-dollar government contracts. This move signals how quickly the AI gold rush is moving from labs to Wall Street.

In a strategic move that could reshape the mobile AI landscape, Apple has secured full access to Google's powerful Gemini model. The tech giant plans to use this access to accelerate development of its own lightweight AI systems through an advanced technique called model distillation. While maintaining its independent research path, Apple aims to blend Gemini's capabilities with its own innovations - potentially unveiling the results at this year's WWDC. This collaboration signals a shift in the AI arms race from raw computing power to smarter training methods.

As the AI race heats up, OpenAI is making strategic cuts to concentrate its firepower. The company is pulling resources from scattered experimental projects to focus squarely on two lucrative fronts: programming tools for developers and premium services for businesses. This pivot comes as rival Anthropic eyes an IPO, putting pressure on OpenAI to demonstrate clear paths to profitability in an increasingly competitive market.

Google's Gemini AI assistant just made switching platforms much easier with its new Memory Import feature. No more starting from scratch – users can now seamlessly transfer their conversation history and preferences from other AI tools. The update offers two migration options: a quick summary transfer or a full data import for those who want to preserve extensive interaction histories. This move not only reduces user frustration but also signals a shift toward more open data practices in the competitive AI assistant market.

Despite reporting a $3.4 billion loss in 2025, Meituan is doubling down on AI development with its new LongCat language model and Xiaotuan assistant. The food delivery giant aims to transform how users interact with local services through smarter search and task automation. While short-term profits suffer, CEO Wang Xing believes AI integration will redefine the $50 billion local services market.

Google has rolled out its 'Search Live' feature globally, transforming how we interact with our surroundings through AI. By simply pointing your phone's camera at objects and asking questions, you'll get instant voice answers powered by the new Gemini 3.1 Flash Live model. This move comes as competitors like OpenAI and Luma AI challenge Google's dominance in AI-powered search.