AI News
Latest updates from the AI world — model releases, funding, research & regulation.
Anthropic Raises $4 Billion in Series E Funding Round
AI safety startup Anthropic has closed a $4 billion Series E funding round, bringing its total valuation to over $15 billion. The round was led by Google with participation from Amazon and other strategic investors. The fresh capital will be used to accelerate the development of Claude models and expand Anthropic's research into AI safety. This makes Anthropic one of the best-funded AI companies in the world, second only to OpenAI.
Google Releases Gemini 2.0 Ultra with Multimodal Capabilities
Google has launched Gemini 2.0 Ultra, a significant upgrade to its flagship AI model. The new version features enhanced multimodal understanding, improved reasoning, and better performance on coding benchmarks. Gemini 2.0 Ultra is now available to Google One AI Premium subscribers and through the Gemini API. Google claims the model outperforms GPT-4 on several key benchmarks.
OpenAI Launches GPT-5 with Major Reasoning Improvements
OpenAI has officially released GPT-5, its most capable model to date. The new model demonstrates significant improvements in complex reasoning, mathematics, and coding tasks. GPT-5 achieves state-of-the-art scores on multiple benchmarks, outperforming previous models by a substantial margin. The release comes with both API access for developers and an updated ChatGPT experience for consumers. OpenAI CEO Sam Altman described it as a major step toward more capable and reliable AI systems.
Anthropic Raises $4 Billion in Series E Funding Round
AI safety startup Anthropic has closed a $4 billion Series E funding round, bringing its total valuation to over $15 billion. The round was led by Google with participation from Amazon and other strategic investors. The fresh capital will be used to accelerate the development of Claude models and expand Anthropic's research into AI safety. This makes Anthropic one of the best-funded AI companies in the world, second only to OpenAI.
Google Releases Gemini 2.0 Ultra with Multimodal Capabilities
Google has launched Gemini 2.0 Ultra, a significant upgrade to its flagship AI model. The new version features enhanced multimodal understanding, improved reasoning, and better performance on coding benchmarks. Gemini 2.0 Ultra is now available to Google One AI Premium subscribers and through the Gemini API. Google claims the model outperforms GPT-4 on several key benchmarks.
EU AI Act Enforcement Begins: What It Means for AI Companies
The European Union has officially begun enforcing the AI Act, the world's first comprehensive AI regulation framework. Companies using high-risk AI systems must now comply with transparency requirements, safety assessments, and documentation standards. Non-compliance can result in fines of up to 35 million euros or 7% of global turnover. The regulation affects AI companies operating in or selling to EU markets.
Meta Open-Sources Llama 4 with 400B Parameters
Meta has released Llama 4, its latest open-source large language model with 400 billion parameters. The model is available for free download on Hugging Face and can be used for commercial applications. Llama 4 reportedly matches or exceeds GPT-4 performance on many benchmarks while being fully open-source. This release continues Meta's strategy of open-sourcing frontier models to accelerate the AI ecosystem.
Stability AI Releases New Image Generation Model
Stability AI has launched its latest image generation model, featuring improved photorealism, better text rendering in images, and enhanced control over artistic styles. The new model is available through the Stability AI API and DreamStudio platform. It represents a significant step forward in accessible, high-quality image generation for creators and developers.
Microsoft Copilot Now Deeply Integrated into Windows 12
Microsoft has announced deep integration of Copilot AI into Windows 12, making it the first operating system with native AI assistant capabilities built at the system level. Users can summon Copilot from any application, search across all files and apps, and use it for system-level tasks. The integration is powered by GPT-4 and runs on-device for privacy-sensitive tasks.
The State of AI Agents in 2025: What's Working and What's Not
AI agents — systems that autonomously plan and execute multi-step tasks — have gone from experimental to practical in 2025. Companies like Anthropic, OpenAI, and startups like Cursor and Replit have shipped agent-powered products used by millions. This analysis covers the current state of AI agents, common failure modes, breakthrough use cases, and where the technology is headed in the next 12 months.
Perplexity AI Reaches $9 Billion Valuation After Latest Round
AI search startup Perplexity has raised a new funding round valuing the company at $9 billion. The company has seen explosive growth, serving 200 million monthly active users. Perplexity plans to use the funds to expand its API business, launch enterprise products, and compete more directly with Google Search. The round was led by institutional investors with participation from several tech executives.
Anthropic Launches Claude for Agentic Workflows
Anthropic has released new capabilities for Claude designed specifically for agentic use cases — tasks where the AI must plan, execute, and iterate across multiple steps. The new features include improved tool use, better long-context reasoning, and a new computer use capability that lets Claude interact with web browsers and desktop applications. Enterprise customers can now deploy Claude agents for automated research, data analysis, and workflow automation.