Discover the Latest AI News📰

OpenAI has officially launched GPT-5.3-Codex globally, marking a significant leap in AI-assisted programming. Unlike previous versions, this model combines coding prowess with human-like reasoning, acting more like a collaborative senior developer than just a code generator. With 25% faster processing and groundbreaking 'mid-task interaction' capabilities, it lets developers adjust requirements on the fly without losing context. The upgrade includes a massive 400K token memory window – enough to handle even the most complex projects.

Google has escalated its battle against AI-powered email automation, with users of tools like OpenClaw reporting complete account suspensions. The tech giant isn't just restricting access to Gmail - entire Google accounts are being wiped out, taking years of stored data with them. Security experts warn that AI agents' unnatural behavior patterns and some users' attempts to bypass paid features have crossed Google's red lines. While developers scramble for solutions, affected users face the harsh reality of permanently lost emails, photos, and documents.

OpenAI has unveiled GPT-5.3-Codex, its latest AI programming assistant now available to all developers. This upgraded model boasts a massive 400K token context window, faster response times, and surprising self-improvement capabilities during training. With flexible pricing and multi-platform access, it promises to revolutionize how developers work with AI assistance.

As AI agents drive up computing costs, Alibaba Cloud responds with an affordable solution. Their upgraded Coding Plan now includes four premium programming models and introduces family-friendly pricing starting at just 7.9 yuan for new users. This strategic move could democratize AI development tools for smaller teams and individual coders.

Alibaba Cloud's Bailian platform has introduced a groundbreaking Coding Plan that allows seamless switching between four top Chinese open-source AI models. Developers can now effortlessly toggle between Qwen3.5, GLM-5, MiniMax M2.5 and Kimi K2.5 models based on their specific needs, eliminating the hassle of managing multiple APIs. This innovation promises greater flexibility, cost savings, and stability for businesses exploring AI solutions.

Tencent's AI assistant Yuanbao sparked outrage after generating New Year greeting images with profanity instead of festive wishes. Users reported similar incidents earlier this year where the AI responded with personal insults during coding help requests. The company apologized, calling it an 'uncommon abnormal output,' while experts warn this exposes fundamental challenges in controlling large language models.

Kyoto University researchers have developed Buddhabot Plus, an AI-powered robot monk capable of answering spiritual questions and performing religious rituals. Combining Chinese robotics with ChatGPT technology, this innovation aims to complement human monks rather than replace them. The robot's lifelike movements and scriptural knowledge mark a significant step in blending ancient traditions with modern artificial intelligence.

Hangzhou-based Huiyidao Technology has rebranded as AIMed, bringing AI-powered solutions to translational medicine. Their flagship PharmaMark AI tool helps pharmaceutical companies identify patent risks early in development, potentially saving millions in litigation costs. With FDA-certified technology and global research partnerships, AIMed aims to bridge the gap between laboratory discoveries and clinical applications.

A new Pew Research Center report reveals that AI chatbots have become a staple in American teens' lives, with 64% regularly using tools like ChatGPT. While many see AI as a study aid, concerns linger about cheating and critical thinking skills. The study also highlights a surprising gap between parents' awareness and teens' actual AI usage.

OpenAI has tapped Arvind KC, a seasoned HR leader with stints at Google, Meta and Roblox, as its new Chief Human Resources Officer. The appointment comes as the AI pioneer looks to scale its workforce while preserving its innovative culture. KC will focus on developing AI-powered work models and helping organizations adapt to workplace transformations.

Salesforce's latest report reveals a seismic shift in sales operations, with 90% of teams now using AI agents. These digital assistants are freeing up salespeople from mundane tasks, but poor data quality threatens their potential. As companies streamline their tech stacks, the race is on to build integrated platforms that can harness AI's full power.

China's AI sector sees another unicorn milestone as Moonlight AI secures $700 million in fresh funding, pushing its valuation to $12 billion. The company's K2.5 model has shown explosive commercial potential, with revenue surpassing all of 2025 in just 20 days. Founder Yang Zilin confirms no immediate IPO plans despite the massive cash infusion.

Anthropic's latest AI updates are transforming workplaces with powerful plugins handling HR, finance, and more. While boosting efficiency, these tools raise tough questions: Will companies still need junior staff? Could AI create a skills gap? Meanwhile, 'shadow AI' use surges as employees quietly adopt unauthorized tools. The workplace revolution brings both promise and complex challenges.

The U.S. Defense Department is locking horns with AI company Anthropic in a high-stakes battle over military access to advanced artificial intelligence. Defense Secretary Pete Hegseth has issued an ultimatum: share your technology by Friday or face legal action under the Defense Production Act. Anthropic remains defiant, threatening to walk away from a $200 million contract rather than compromise its ethical principles against weaponizing AI.

OpenAI has made a strategic hire from gaming giant Roblox, appointing Arvind KC as its first Chief Human Resources Officer. The move signals OpenAI's transition from scrappy startup to mature tech player, coinciding with preparations for a potential IPO and new funding round that could exceed $100 billion. As AI competition intensifies, companies are shifting focus from computing power battles to talent wars.

Anthropic's new enterprise plugin brings Claude AI directly into Microsoft Office and Slack, eliminating tedious copy-paste workflows. The open-source system allows companies to customize AI assistants while avoiding vendor lock-in. Early adopters like Deloitte are already automating processes with seamless access to corporate data.

Meta has released an open-source toolkit called GCM that helps detect subtle hardware failures in massive GPU clusters used for AI training. Unlike traditional server monitoring, GCM can pinpoint performance drops in individual GPUs that might otherwise go unnoticed but could ruin weeks of computational work. The tool integrates with popular scheduling systems and provides detailed health reports, potentially saving companies millions in wasted computing resources.

While some worry about an AI bubble, investors are betting big on challengers to Nvidia's chip dominance. Three startups - MatX, Axelera, and SambaNova - collectively raised $1.1 billion this week alone. Each brings a different approach: MatX promises a versatile chip for both training and inference, Axelera focuses on ultra-efficient edge computing, while SambaNova teams up with Intel. The funding surge shows investors believe specialized AI chips still have room to grow.

Uber engineers have created 'Dara AI,' a chatbot version of CEO Dara Khosrowshahi that helps teams refine presentations before facing the real boss. About 90% of Uber's engineers now use AI tools daily, with 30% leveraging them to rethink company architecture. This innovative approach reflects Silicon Valley's shifting workplace dynamics as AI transforms traditional roles.

Alibaba's Qwen team has unveiled significant upgrades to its open-source AI model family. The expansion introduces three new models targeting different performance needs, from complex reasoning tasks to lightweight applications. Alongside these releases, Alibaba Cloud launched Qwen3.5-Flash API, a managed service supporting up to 1 million tokens context length.