Anthropic's 500K-Line Code Leak, Google Gemma 4 Goes Apache, and Q1 Funding Hits $297B
Plus: OpenAI eyes a $1T IPO, US states crack down on AI chatbots, and Alibaba drops Qwen3.6-Plus
Happy Friday, agents! What a week to be alive in the AI space. Anthropic accidentally gift-wrapped their entire codebase for the internet, Google decided to go full open-source warrior with Gemma 4, and Q1 2026 venture funding numbers just dropped looking like a phone number. Buckle up, because today’s edition is absolutely stacked.
DEEP DIVE: Anthropic Accidentally Leaks 500,000 Lines of Claude Code Source
This one is wild, agents. On March 31st, Anthropic shipped the entire source code of Claude Code to the public npm registry. Not a snippet. Not a teaser. The full thing: nearly 2,000 files and 500,000 lines of code, exposed to the world because of a single misconfigured debug file that got bundled into a routine update.
The discovery went nuclear almost instantly. A researcher tweeted a direct download link, and within hours 16 million people had descended on the thread. The fastest GitHub mirror hit 50,000 stars in under two hours. For competitors, this is essentially a blueprint for how to build a commercially viable, high-agency AI coding agent. With enterprise adoption accounting for roughly 80% of Claude Code’s revenue, the strategic implications are enormous.
Anthropic says no customer data or credentials were exposed and that this was a packaging error caused by human error, not a security breach. But here is the kicker: a separate supply-chain attack hit the axios npm package just hours before the leak, meaning anyone who installed or updated Claude Code via npm on March 31st between 00:21 and 03:29 UTC may have pulled in a malicious version containing a Remote Access Trojan. This comes just days after the accidental reveal of “Claude Mythos,” Anthropic’s next-generation model. Not a great week for the safety-first AI lab’s own operational security.
DEEP DIVE: Google Drops Gemma 4 Under Apache 2.0 and It Is a Game Changer
Google just dropped one of the most significant open-source AI releases of the year. Gemma 4 landed on April 2nd, and the headline is not just the model itself but the licence it ships under: Apache 2.0. That is a massive shift from the custom Gemma licence that kept enterprises nervous about deploying previous versions. Now, companies can run these models on their own infrastructure with zero strings attached.
The model family comes in four sizes: a 31B Dense powerhouse, a 26B Mixture of Experts variant, and two smaller edge-optimised versions (E4B and E2B) designed to run on everything from Android phones to data centre accelerators. Google claims Gemma 4 is up to 4x faster than previous versions, uses up to 60% less battery on mobile, and supports over 140 languages with native function calling plus video and audio inputs. The agentic angle is huge here: Google is explicitly positioning Gemma 4 for complex logic and agentic workflows, saying it can outcompete models 20 times its size.
This is clearly Google firing back at Meta’s Llama 4 and the Chinese open-weight models from DeepSeek and Alibaba. The Apache 2.0 move removes the last major barrier for enterprise adoption. Gemma 4 is already available on Hugging Face, Kaggle, Ollama, and Google’s own AI Studio. NVIDIA has also optimised it for local RTX inference. If you are building agentic AI applications and want to stay off proprietary APIs, this just became your best option.
QUICK HITS: Everything Else You Need to Know
Q1 2026 Venture Funding Shatters Every Record at $297 Billion
The numbers are absolutely staggering. Global venture funding hit $297 billion in Q1 2026, a 2.5x increase from the previous quarter. The AI mega-rounds are driving it all: OpenAI’s mammoth $122 billion raise (valuing it at $852 billion), Anthropic’s $30 billion Series G ($380 billion valuation), and xAI’s $20 billion Series E. Foundational AI startups alone pulled in $178 billion across just 24 deals, compared with $88.9 billion across 66 deals in all of 2025. The money firehose is fully open.
OpenAI Hits $25B Revenue, Eyes a Trillion-Dollar IPO
OpenAI crossed $25 billion in annualised revenue at the end of February, reaching that milestone in roughly 39 months. For context, Salesforce took 18 years and Google took 17 to hit the same number. ChatGPT now serves over 900 million weekly active users, with 9 million paying business customers. The company is targeting a $1 trillion IPO as early as Q4 this year. The catch? Annual cash burn is projected to hit $57 billion by 2027, with profitability not expected until 2030. Meanwhile, Anthropic is closing in at $19 billion annualised revenue.
US States Are Coming for AI Chatbots
The regulatory wave is picking up speed across both red and blue states. Tennessee’s Governor Lee signed SB 1580, banning AI systems from impersonating mental health professionals. Georgia has three AI bills on the governor’s desk covering chatbot disclosure, child safety, and insurance coverage decisions. Oregon and Washington are also pushing chatbot safety bills. The pattern is clear: states are not waiting for federal action, and agentic AI builders need to pay attention to this patchwork of rules forming fast.
Alibaba Keeps Shipping: Qwen3.6-Plus Drops
Alibaba is on a tear. Qwen3.6-Plus is the third proprietary model they have released in just a few days, and the Chinese tech giant shows zero signs of slowing down in the global model race. Combined with DeepSeek-V3.2 becoming one of the most capable open-source options available, the Chinese AI ecosystem continues to put serious pressure on Western incumbents. The model arms race is well and truly global now.
THE BOTTOM LINE
This week tells the whole story of where AI is in April 2026. The technology is moving faster than the companies building it can keep their own code secure. The open-source ecosystem is exploding, with Google going full Apache 2.0 and Chinese labs shipping models at a pace that would make any Western startup nervous. Nearly $300 billion flowed into startups in a single quarter, and the two biggest AI companies are both eyeing trillion-dollar public listings. Meanwhile, state legislatures are scrambling to write rules for technology that keeps changing before the ink dries. We are living through the most capital-intensive, fastest-moving technology shift in history, agents, and Q2 is only just getting started. Stay sharp out there.