openais-gpt-5-leak-new-ai-laws-and-the-future-of-coding-tools-img-0

AI_Distilled #106: What’s New in AI This Week

Become an AI Generalist that makes $100K (in 16 hours)

openais-gpt-5-leak-new-ai-laws-and-the-future-of-coding-tools-img-1

One of the biggest IT giants, TCS laid off 12,000 people this week. And this is just the beginning of the blood bath.

In the coming days you’ll see not thousands, but millions of more layoffs & displacement of jobs. So what should you do right now to avoid getting affected?

Invest your time in learning about AI. The tools, the use cases, the workflows – as much as you can.

Join the World’s First 16-Hour LIVE AI Upskilling Sprint for professionals, founders, consultants & business owners like you.

Date: Saturday and Sunday, 10 AM - 7 PM.

REGISTER NOW (ONLY 500 FREE SEATS)

Rated 4.9/5 by global learners – this will truly make you an AI Generalist that can build, solve & work on anything with AI.

In just 16 hours & 5 sessions, you will:
✅ Learn how AI really works by learning 10+ AI tools, LLM models and their practical use cases.
✅ Learn to build and ship products faster, in days instead of months
✅ Build AI Agents that handle your repetitive work and free up 20+ hours weekly
✅ Create professional images and videos for your business, social media, and marketing campaigns.
✅ Turn these AI skills into10$k income by consulting or starting your own AI services business.

All by global experts from companies like Amazon, Microsoft, SamurAI and more. And it’s ALL. FOR. FREE. 🤯 🚀

$5100+ worth of AI tools across 2 days — Day 1: 3000+ Prompt Bible, Day 2: Roadmap to make $10K/month with AI, additional bonus: Your Personal AI Toolkit Builder.

This week witnessed major developments in the AI world, from accidental leaks and new open-source models to a growing rivalry between tech giants. We have curated the most significant stories you need to know.

So, let’s get started!

LLM Expert Insights,

Packt

📈LATEST DEVELOPMENT

OpenAI Releases Open-Weight Models

OpenAI has launched two new open-weight models, gpt-oss-120b and gpt-oss-20b. These models are designed for advanced reasoning, can be customized, and are available under a permissive Apache 2.0 license. The models feature powerful instruction-following and tool use, with a focus on safety through comprehensive testing. They can be run on various devices, from data centers to laptops.

GitHub Accidentally Leaks OpenAI's GPT-5

GitHub briefly and accidentally published a blog post announcing OpenAI's new GPT-5 models, which have since been removed. The leak reveals four new variants with "major improvements in reasoning, code quality, and user experience." GPT-5 will have "enhanced agentic capabilities" and can handle complex coding tasks with minimal prompting. A live stream from OpenAI is expected to confirm the details.

Meanwhile, OpenAI is teasing a "LIVE5TREAM" for this Thursday, hinting at the launch of its new GPT-5 model. Recent social media posts from OpenAI's CEO also suggest a reveal is imminent.

EU's New AI Act Rules for General-Purpose AI Are Now Live

The European Union has activated the AI Act's obligations for providers of general-purpose AI (GPAI) models. These new rules, which came into effect on August 2, 2025, aim to improve transparency, safety, and accountability. Providers must now offer clearer information on model training and adhere to copyright protections. The most advanced models, those with systemic risks, face additional safety and security requirements. A Code of Practice will also help providers ensure compliance.

Kaggle Launches Game Arena for AI Evaluation

Google DeepMind and Kaggle have introduced Game Arena, a new open-source platform for evaluating AI models. This platform allows models to compete head-to-head in strategic games like chess, providing a transparent and verifiable measure of their problem-solving and reasoning skills. The method addresses the limitations of traditional benchmarks, such as memorization and subjectivity, with an all-play-all ranking system.

OpenAI to Build European Data Centre in Norway

OpenAI is partnering with Aker ASA and Nscale to build its first European "Stargate" data center in Norway, a $1 billion project. Powered by local hydropower, the facility will initially house 100,000 Nvidia processors and could expand tenfold. The initiative, crucial for AI development, provides infrastructure for developers, researchers, and startups across Europe.

Anthropic Cuts Off OpenAI's API Access to Claude

Anthropic has revoked OpenAI's API access to its Claude models, citing a violation of its terms of service. According to sources, OpenAI was using the access for benchmarking and to test Claude's capabilities in areas like coding ahead of the GPT-5 launch. OpenAI called the move disappointing, noting its own API remains open to Anthropic.

📈HELPFUL AI TOOLS

In this issue of AI Distilled, we've curated a selection of AI-powered coding tools.
Have you explored them all, or are a few still new to you?

Now’s the perfect time to vibe code: go check them out!

Cursor

Features and Highlights:
AI-native IDE (fork of VS Code), context-aware chat, auto-fix, agent mode, and multiline suggestions.

Ideal For: Professional developers, indie hackers

Webpage: https://cursor.com/

GitHub Copilot

Features and Highlights:
Code suggestions, GitHub-native workflows, IP-safe for enterprises.

Ideal For: Developers, software teams, enterprises

Webpage: https://github.com/features/copilot

Replit

Features and Highlights: In-browser IDE with AI chat, live collaboration, prompt-based app building.

Ideal For: Beginners, educators, startup teams

Webpage: https://replit.com/

Lovable AI

Features and Highlights: MVP builder from prompts; great for non-coders to test startup ideas.

Ideal For: Startup founders, product managers, non-tech users

Webpage: https://lovable.dev/

Bolt

Features and Highlights: Fast app prototyping from plain English, minimalist interface.

Ideal For: No-code/low-code users, indie founders

Webpage: https://bolt.new/

Features and Highlights: Generates production-grade code with visual transparency.

Ideal For: Frontend developers, UI designers

Webpage: https://v0.dev/

Claude

Features and Highlights: LLM used behind the scenes for code generation and comprehension.

Ideal For: Toolmakers & app builders integrating AI

Webpage: https://www.anthropic.com/claude-code

Windsurf

Features and Highlights: Predictive code completion, OpenAI-backed, IDE plugins.

Ideal For: Backend/frontend developers

Webpage: https://windsurf.com/

EXPERT INSIGHTS

From the Packt Bookshelf: The LLM Engineer’s Handbook by Paul Iusztin and Maxime Labonne

Before you add the fancy optimizations like hybrid search, multi-hop reasoning, adaptive prompts, you need to master the foundation of Retrieval-Augmented Generation (RAG). In this excerpt from The LLM Engineer’s Handbook, Paul Iusztin and Maxime Labonne break down “vanilla” RAG into its three essential modules:

- Ingestion: Extract, clean, chunk, and embed your data before loading it into a vector DB.
- Retrieval: Embed the user’s query, find the most relevant chunks using similarity search, and return the right context.
- Generation: Combine the query with retrieved context in a prompt template, then let your LLM craft the answer.

Once these fundamentals click, you’ll know exactly how to layer in more advanced techniques—without losing what makes the system accurate and reliable.

Read the full article on Substack →

Built something cool? Tell us.

Whether it's a scrappy prototype or a production-grade agent, we want to hear how you're putting generative AI to work. Drop us your story at [email protected] or reply to this email, and you could get featured in an upcoming issue of AI_Distilled.

📢 If your company is interested in reaching an audience of developers and, technical professionals, and decision makers, you may want toadvertise with us.

If you have any comments or feedback, just reply back to this email.

Thanks for reading and have a great day!