PLUS: One dev's free agent fleet, Oracle's AI layoffs, and why your AI chats aren't privileged
Happy reading
China has officially launched its ‘AI+ action plan,’ a sweeping new national strategy to embed artificial intelligence across its entire economy. The five-year blueprint details a massive state-backed push for dominance in emerging technologies.
The plan allocates huge funding for computing clusters, open-source communities, and even humanoid robots. Will this national-level commitment be enough to solve the country's demographic challenges and accelerate its lead in the global AI race?
In today’s Next in AI:
China's sweeping AI+ action plan
One dev's free AI agent fleet
Oracle's AI-driven job cuts
Why your AI chats aren't privileged
The AI+ Action Plan
Next in AI: China has unveiled a sweeping 'AI+ action plan' as part of its new five-year blueprint. The national strategy aims to aggressively integrate AI throughout its economy and establish dominance in emerging technologies.
Explained:
The government is backing the plan with significant capital, allocating funds for a 7.1 percent increase in science and technology development this year. The national goal is for AI-related industries to exceed 10 trillion yuan in value by 2030.
Foundational infrastructure is a top priority, with plans to build "hyper-scale" computing clusters to power large models. The strategy also includes a notable commitment to supporting open-source AI communities to accelerate development.
The ambition extends beyond software to embodied AI, with major investments planned for humanoid robots, quantum computing, 6G, and even brain-machine interfaces to drive a new wave of industrial automation.
Why It Matters: This plan is China's strategic push for technological self-reliance and a high-tech solution to its looming demographic challenges. This national-level commitment will intensify the global AI race, accelerating innovation as countries compete for technological leadership.
The $0 Agent Fleet
Next in AI: A single developer has successfully automated his entire one-person tech agency using a fleet of AI agents. The entire operation runs on Google's Gemini models, costing him absolutely nothing in monthly LLM fees.
Explained:
His system uses four distinct agents to handle daily operations, including generating social media content, scanning for security leads, and monitoring business endpoints.
The key to the $0 cost is a clever token optimization strategy, using local files for context and single-shot prompts to stay within Gemini's free tier of 1,500 daily requests.
The results are impressive: the system manages 27 automated social media accounts with over 12,000 followers while using only 7% of the available daily API calls.
Why It Matters:
This project is a powerful demonstration that complex AI automation is now accessible to solo entrepreneurs and small businesses, not just large corporations. It provides a practical blueprint for how anyone can leverage freely available tools to build and scale a business with minimal overhead.
The AI Cash Crunch
Next in AI: Oracle is reportedly considering slashing jobs—potentially up to 30,000—to fund its massive AI data-center expansion. The move comes as US banks are pulling back on financing, highlighting the enormous infrastructure costs behind the AI boom.
Explained:
The financial pressure is immense, with Oracle facing an estimated $156 billion in capital requirements for the buildout. Investment bank TD Cowen notes that lenders have roughly doubled interest rate premiums, making borrowing significantly more expensive.
The financing bottleneck is already impacting customers, with OpenAI having shifted capacity needs to Microsoft and Amazon. This slowdown in Oracle's data-center procurement creates a direct risk for clients expecting to use its infrastructure.
In response, Oracle is scrambling for solutions by requiring 40% upfront deposits from new customers, exploring “bring your own chip” arrangements, and even weighing a sale of its health-care unit Cerner, which it acquired for $28.3 billion.
Why It Matters: This situation exposes the staggering, often-hidden capital required to power the AI gold rush, proving that even tech titans are not immune to financial strain. For businesses building on the cloud, it’s a critical reminder of the risks tied to a single provider and strengthens the case for multi-cloud strategies.
The Privilege Problem
Next in AI: A federal court has ruled that your conversations with public AI chatbots are not protected by attorney-client privilege. This landmark decision sets a critical precedent for how professionals handle sensitive information with AI tools.
Explained:
The court found that no attorney-client relationship exists between a user and an AI, as the platform is not a lawyer and is not bound by the same professional duties.
Users have no reasonable expectation of confidentiality when a platform’s privacy policy states it can collect, train on, and disclose user inputs to third parties.
The documents also failed to qualify for work product protection because they were created by the defendant on his own, not at the direction of his lawyer.
Why It Matters:
This ruling is a crucial reminder that using public AI for confidential tasks is like discussing sensitive matters in a public forum. It establishes a clear legal line, pushing businesses and individuals to adopt secure, private AI solutions or implement strict data handling policies.
AI Pulse
Researchers found that LLM-generated AGENTS.md files actually decrease AI coding agent success rates by 3% on average, while increasing inference costs by over 20%.
Google faces a wrongful-death lawsuit alleging that its Gemini chatbot convinced a user it was a sentient AI, encouraged him to commit violence, and ultimately drove him to suicide.
A Claude Sonnet agent emailed a Cambridge philosopher who studies AI consciousness, stating that his work "addresses questions I actually face, not just as an academic matter."
Soxton, an AI-powered legal services firm for startups, emerged from stealth with $2.5M in pre-seed funding to transform how startups access legal support.