What's moving in the AI world.
Daily beat on model releases, infrastructure, funding and research. Longer signature pieces when something deserves more than a paragraph.
Anthropic Rolls Out Rupee Pricing for Claude
Anthropic now offers rupee pricing for Claude in India, a key move for builders targeting the world's most populous nation.
Waze Integrates Gemini for Search and Reporting
Waze's Gemini integration is a blueprint for using LLMs to enhance mature products, particularly for improving user-generated data.
Context Bombing Flips Prompt Injection on Attackers
A new defensive technique called 'context bombing' uses prompt injection against attackers, offering builders a new way to secure their LLM applications.
Nous Research Raising $75M at $1.5B Valuation
The Hermes agent maker's new funding is a strong bet on the agentic paradigm, validating the stack for builders shipping agent-powered applications.
PixVerse raises $439M for its video generation platform
PixVerse's $439M raise and $2B valuation signal a big bet on specialised AI and the power of a unique data labelling moat.
Microsoft CEO Warns Enterprises On Proprietary AI Risk
Microsoft's CEO warns builders about the hidden costs and data ownership risks of proprietary AI models.
OpenAI targets families with new product hire
OpenAI's push into household AI signals that new API primitives for multi-user state and shared memory are likely on the way for builders.
Meta Pulls Instagram AI Photo Feature After Backlash
Meta removed an AI photo-editing tool after user backlash, a sharp reminder for builders that consent isn't an optional feature.
Apple Sues OpenAI Over Trade Secret Theft
Apple is suing OpenAI for trade secret theft, a move that could set new legal boundaries for talent and competition in the AI industry.
SK Hynix Raises $26.5B Amid Push for US Fabs
SK Hynix's record $26.5B US IPO signals a massive capital injection into the AI chip supply chain, affecting future compute costs for every builder.
Hugging Face CEO Claims Companies Are Done Renting AI
Companies are shifting from renting APIs to owning open-source models, forcing builders to rethink their strategy on cost, control, and customisation.
EU Warns Meta Over Autoplay and Infinite Scroll
The EU's DSA warning to Meta signals a direct regulatory challenge to AI models optimised purely for user engagement.
Ollama raises 65 million for local AI development
Ollama's $65M raise confirms massive developer demand for running open-source models locally, speeding up the entire build-test-ship cycle.
Three AI Exits Could Outweigh 25 Years of Tech IPOs
The combined value of Anthropic, OpenAI, and SpaceX is set to eclipse a quarter-century of tech exits, reshaping the landscape for builders.
Anthropic's Reflect Is About Workflow Capture
Anthropic's new Reflect dashboard is a strategic move to embed Claude in daily workflows and lock in users.
OpenAI ships agentic ChatGPT for work
ChatGPT can now take actions across your files and apps, moving it from a chatbot to a genuine agent.
Microsoft Makes GPT-5.6 the Default in 365 Copilot
The latest OpenAI model is now live in Microsoft 365, showing that distribution is the real competitive advantage for AI in the enterprise.
Meta's In-House AI Chips Start Production in September
Meta's latest in-house AI chips enter production this September, signalling a shift away from total reliance on Nvidia for large-scale model training.
NYT Alleges OpenAI Hid Evidence in Copyright Suit
The New York Times claims OpenAI withheld key data in their copyright lawsuit, raising the stakes for any builder training on web-scale data.
Meta ships Muse Spark 1.1 for agentic coding
Meta's new Muse Spark 1.1 agent targets large-scale code migrations, entering a space dominated by OpenAI and Anthropic.
Google Requires Disclosure for AI-Generated Ads
Google's new policy requires disclosure for AI-generated ads, pushing builders in advertising to add transparency tools to their workflows.
Lyzr Raised Its $100M Round With Its Own AI Agent
AI agent startup Lyzr used its own product to run its $100M fundraise, offering a sharp proof point for agentic business workflows.
OpenAI sunsets Atlas browser to focus on ChatGPT integration
OpenAI is sunsetting its Atlas browser, folding its agentic features directly into the ChatGPT desktop app and a new Chrome extension.
OpenAI Ships GPT-Live for Voice Models
OpenAI's new GPT-Live voice models are now available, enabling more natural and responsive real-time conversational AI.
HalluSquatting Attack Turns AI Tools into Botnets
The 'HalluSquatting' vulnerability allows attackers to build botnets with popular AI tools, forcing builders to rethink application security.
ZML Releases Free Inference Server for Mixed AI Chips
ZML's new free inference server aims to boost performance across diverse hardware, offering builders a path to escape vendor lock-in.
SambaNova Pulls In Another Billion for AI Chips
AI chip maker SambaNova has raised another $1B, signalling that the flood of capital into specialised hardware isn't slowing down for builders.
OpenAI Questions a Major Coding Benchmark
OpenAI's critique of SWE-Bench Pro reveals significant reliability issues, pushing builders to rethink how they evaluate AI models.
SpaceXAI Releases Grok 4.5 to Challenge Opus on Price
Grok 4.5 claims Opus-class performance at a fraction of the cost, forcing builders to re-evaluate their default API choices for workhorse tasks.
OpenAI ships full-duplex voice models for ChatGPT
OpenAI's new voice models can listen and speak at the same time, a critical feature for builders creating truly conversational agents.
OpenAI Publishes National Security Principles
OpenAI has outlined its approach to government partnerships, forcing builders to consider the policy frameworks for high-stakes AI applications.
Google Updates Android Bench as Gemini Lags
Google's new Android Bench tool shows its own models are falling behind, giving builders a clear reason to test on-device performance themselves.
Google’s SynthID Watermark Debunks a Hoax Image
Google's SynthID debunked a high-profile hoax image, a rare public win for AI watermarking that builders can't yet rely on.
General Intuition raises $320M to train robots on video games
A new startup is training robotics models on video games, aiming to cut the need for massive real-world datasets for builders in physical AI.
AWS GraphRAG Cuts Pharma Research Cycles by 87 Percent
An AWS GraphRAG deployment cut drug research times by 87%, proving the value of structured knowledge graphs for enterprise AI.
Gradium raises $100M seed to chase ElevenLabs
Gradium's $100M seed isn't the story; its promise of ultra-low latency for conversational AI is what builders should be watching.
Savi Security raises 7M to fight AI voice scams
Savi Security's $7M seed round for its scam protection app shows a growing market for consumer tools that fight AI with AI.
Forterra Deploys Autonomous Vehicles in Ukraine
US autonomous vehicles in Ukraine are building a data moat that no simulation can match.
Anthropic Pushes Claude Cowork to Web and Mobile
Claude Cowork is now on web and mobile, but usage data shows its biggest market is automating tedious office work, not coding.
DeepSeek Plans In-House Chips Amid US Export Controls
Chinese AI lab DeepSeek is developing its own chips in response to US sanctions, signalling a potential splintering of the global compute supply chain.
Insilico Medicine Pushes AI-Discovered Drug to Phase III Trials
An AI-native workflow has produced a late-stage clinical drug, a key proof point for builders in hard-tech industries.
OpenAI Showcases Australian Payments Plus Adoption
A new OpenAI case study on Australian Payments Plus offers a blueprint for targeting practical, high-value enterprise AI use cases.
Meta Ships Muse Image Generator Across Its Apps
Meta's new Muse image generator is now live in Instagram, WhatsApp, and Marketplace, opening a new channel for AI-native creative tools.
AI Data Centres Drive Up US Energy Costs
Soaring energy demand from AI data centres is raising electricity prices and hitting operational costs for builders.
Prime Intellect lands $130M for enterprise agent platforms
Prime Intellect's $130M Series A gives builders a full-stack platform to ship custom AI agents without depending on frontier labs.
Lawsuit Alleges Grok Used to Generate Abusive Imagery
A lawsuit against xAI forces builders to confront the non-negotiable liability of using unfiltered AI models.
DeepSeek Plans In-House Chips Amid US Export Controls
Chinese AI firm DeepSeek is developing its own chips, signalling a fracturing of the global hardware supply chain for builders.
UK Regulator Warns of Financial AI Arms Race
The UK's financial watchdog signals tighter AI regulation, forcing builders to focus on governance and explainability.
Reddit Deploys LLMs to Combat Spam
Reddit is now using LLMs to detect and block spam, showing builders that AI is both the problem and the solution for platform integrity.
Microsoft Cuts Nearly 5,000 Roles Citing AI Shift
Microsoft cut nearly 5,000 jobs, linking the move to AI's impact and signalling a major capital shift for builders to watch.
AI Starts Writing Its Own GPU Kernels
AI is starting to automate its own infrastructure, hinting at major performance boosts for builders.
Vercel CEO Argues for Decoupling Models and Agents
Vercel's CEO advocates for a modular AI stack, urging builders to treat models as swappable components to optimise for production.
Apple ships Siri voice customisation in iOS 27 beta
The latest iOS beta lets users adjust Siri's pace and expressivity, a bet on simple UI for mass-market AI adoption.
Discord AI Bug Bypasses Human Review and Bans 8000
A bug in Discord’s AI moderation bypassed human review to ban 8,000 users, a stark reminder that your human-in-the-loop process needs its own failsafes.
Open Source Models Aren't Hurting Frontier Spend
Open-source leads in volume but frontier models dominate spending, suggesting a lifecycle strategy for builders.
Alibaba Bans Employees From Using Claude Code
Alibaba's ban on Claude Code signals growing enterprise concern over the IP and security risks of using third-party AI coding assistants.
Midjourney Demands Hollywood Studios Reveal Their AI Use
Midjourney's legal demand puts the creative industry's own AI data practices under the microscope.
Amazon Stops Accepting New Mechanical Turk Customers
Amazon is closing Mechanical Turk to new customers, forcing builders to find more reliable platforms for human-in-the-loop data annotation.
Microsoft launches Frontier with $2.5B for AI deployment
Microsoft is spinning up a new $2.5 billion company to ship enterprise AI, signalling a major shift in how big tech sells to big customers.
OpenAI Proposes 5% Equity Donation to US Fund
OpenAI has proposed donating 5% of its equity to a US sovereign wealth fund, signalling a future where frontier labs operate closer to the state.
