Welcome, AI Pioneers!

Search is no longer something you do. It’s something done for you.

Google just flipped the script on how we explore the internet. But it isn’t alone. Open-source II-Agent is quietly raising the bar for AI reliability with powerful, lightweight tools. Apple plans to hand developers its secret AI recipe, pushing smarter apps to your fingertips. Meanwhile, Tencent’s AI game engine is turning months of creative work into minutes.

When machines become the primary readers of the internet, who are websites really writing for?

In today’s Generative AI Newsletter:

• Google’s AI agents reinvent web search and browsing
• II-Agent sets a new standard for open-source AI frameworks
• Apple Intelligence opens up to third-party developers
• Tencent launches a game-changing AI engine

Special highlight from our network

The Biggest AI Risks Are the Ones You Can’t See.

Underneath every GenAI interaction lie hallucinations, jailbreaks, and policy violations waiting to happen.

🔐 AI Trust Needs a Scorecard.

Tumeryk’s AI Trust Score™ Manager brings real-time guardrails to your GenAI deployments.

✔️ Monitor & block unsafe AI responses

✔️ Auto-enforce policy-based rules (no PII, jailbreaks, hallucinations)

✔️ Works across OpenAI, Anthropic, Azure & more

✔️ Build your own trust policies — at scale

✔️ Logs, scores & secures — all in real-time

🛡️ Enterprise-ready. Cloud-secure. Compliance-first.

🌐 Google Brings the Web to You with AI Agents

Search as we knew it is over. At I/O 2025, it became clear that Google is reconfiguring the internet around AI agents that don't just retrieve information, but actively navigate, read, and summarize online content on users' behalf.

What’s new:

• AI Mode in Search (US-wide): Every user now gets an agent that reads web pages, answers questions, summarizes content, and handles follow-ups

• Deep Research agent upgrades: It scans dozens of sources to build detailed reports, personalized through Gmail and Drive

• Project Mariner becomes hands-free: Ultra subscribers get agents that click, scroll, and handle 10 tasks in parallel.

• Project Astra integration: Users can now speak to the web via real-time, multimodal agents that see what you see

• Gemini SDK supports agent protocols: Google’s AI stack supports Anthropic’s MCP, helping agents access data across the internet.

This changes not just how people search, but also who visits websites. AI agents, rather than human users, are becoming the primary visitors. This poses a challenge to the ad-supported internet model, which relies on human engagement.

🔓II-Agent: A New Standard for Open-Source AI Agents

Intelligent Internet has released II-Agent, an open-source agent framework that directly challenges the reliability gap in open systems. While most frameworks prioritize modularity, II-Agent focuses on pure task completion and it delivers.

Key facts:

• Claims to beat GenSpark and Manus on benchmarks in just 500 lines of Python.

• Native support for Claude, GPT-4, Gemini, and local open models

• Fully supports tool use, web browsing, and memory

• Designed for production invisibility with local dev simplicity

• Passes 80% of SWE-bench-lite using GPT-4, no fine-tuning

Most open-source agents are either toy demos or dev-time scaffolds. II-Agent flips the script: small enough to audit, strong enough to ship, and fast enough to matter.

🍎 Apple Lets Developers Build on Apple Intelligence

Apple is preparing to open up its Apple Intelligence models to third-party developers, giving them access to the same LLMs powering the company’s native AI features. The announcement will come on June 9 at WWDC, alongside a software development kit.

Key points:

• Developers will get access to the same small on-device models used by Apple

• Larger cloud-based models to be integrated in a later phase

• The SDK is designed to encourage AI-native apps and increase usage of Apple Intelligence

• Apple’s new OS updates will adopt Vision Pro-style interfaces for iPhone, iPad, and Mac

• The revamp includes simpler navigation, redesigned icons, and more unified UX

Most users won’t care about model benchmarks. They’ll care about which phone runs the smartest apps with the least friction. The question is no longer if Apple is late to the party. It’s what happens when the party moves inside the App Store.

🎮 Tencent Launches China’s First AI Game Engine

Tencent just launched Hunyuan Game, a next-gen AIGC engine. Powered by the Hunyuan foundation model, it’s designed to take developers from concept to production-ready visuals in minutes, not months.

What it delivers:

• AI Art Pipeline: Converts inspiration into design and full triviews in a single, AI-tuned flow.
• Real-Time Canvas: Live sketching with instant refinement. AI evolves your draft as you draw.
• 2D Generator: Precise, game-specific generation for UI and concept art with zero misinterpretation.
• Orthographic Character Views: Auto-generates 360° views, ready for rigging, no edits, no waiting.

Tencent is encoding game studio instincts into algorithms. When AI can generate game worlds from a sentence, what happens to the boundary between developer and player? The future of play might not be coded. It could be prompted.

🌟 Help Us Make the Generative AI Daily Newsletter Even Better! 🌟

We’d love to hear from you!
Your opinion really counts and helps us make our newsletter even better for you. We’re always working to make it more useful and interesting.

This super short survey is your chance to tell us what you think. 

🚀 Boost your business with us—advertise where 10M+ AI leaders engage

🌟 Sign up for the first AI Hub in the world.

📲 Our Socials