Hey {{first_name}}
Welcome to The AI Report — the #1 AI newsletter for business leaders.
We believe you don’t need to be technical. Just informed.
Inside: real-world use cases and Partner Perspectives you won’t find anywhere else:
• 1. 💻 Get rid of your personal data from Google with Incogni
• 2. 📖 OpenAI reveals how its Codex coding agent works
• 3. ⏰ How AI saved employees hours daily
• 4. 🦖 Scientists launch app that identifies dinosaur footprints
• 5. 👓 Google adds agentic vision to Gemini 3 Flash
Read Time: 5 minutes
TOGETHER WITH INCOGNI
Remove your personal data from Google search results
Your phone number, address, DOB, even your SSN can show up in a Google search. Marketers buy it, scammers exploit it, and in the worst cases, it ends up for sale on the dark web.
Incogni Unlimited helps you take control. Our privacy agents hunt down and remove your personal data from an unlimited number of sites. With automation across 420+ broker types—including People Search platforms and private databases—Incogni erases your information where others can’t reach.
Protect your identity before it’s too late. Use code AIREPORT for an exclusive 55% discount on unlimited removals.
Latest in AI
OpenAI reveals how its Codex coding agent works
🚨 Our Report
OpenAI engineer Michael Bolin has published a detailed technical breakdown of how the company's Codex CLI coding agent works internally, revealing the "agentic loop" that powers tools capable of writing code, running tests, and fixing bugs with human supervision. The disclosure comes as AI coding agents like Codex with GPT-5.2 and Claude Code with Opus 4.5 are reaching a level of usefulness that some compare to ChatGPT's early impact.
🔓 Key Points
The post explains Codex's core "agent loop," where the model takes user input, generates a response or requests a tool call (like running a shell command), executes it, appends the output, and repeats until the task is complete.
Bolin acknowledges engineering challenges including quadratic prompt growth and cache misses, noting that every request is fully stateless, meaning the entire conversation history is sent with each API call rather than stored server-side.
Unlike ChatGPT, OpenAI has open-sourced its Codex CLI client on GitHub, allowing developers to examine the implementation directly, a level of transparency not extended to its other products.
🔐 Relevance
This breakdown offers a rare glimpse into how agentic tools are built and where their limitations lie. While Codex excels at rapid prototyping and boilerplate code, the post confirms what many developers already suspect: these tools remain brittle beyond their training data and still require human oversight for production-quality work.
Partner Perspective
Seattle NWSL Coach Uses AI to Shape Game Tactics
Partner Column exclusively available in this edition of The AI Report
By Karen Odash — Labor And Employment Attorney at Fisher Phillips
“The question for every club isn’t whether to adopt AI—it’s how quickly you can integrate it into your decision-making before your rivals do.”
The AI Report Podcast
Case Study
How AI saved employees hours daily
A large national wholesaler and distributor of shipping, industrial, and packaging supplies needed to modernize monolithic back-end systems that were hindering growth and scalability.
Pattern: AI-assisted microservices architecture where automated workflows handle data distribution and task management asynchronously, while human engineers oversee deployment, training, and edge cases.
Why it matters: Employees reclaimed significant time daily through asynchronous task management and reliable data access, enabling better inventory control and faster client service.
Metric: Over 25,000 legacy applications decommissioned and 230+ REST microservices created across a 7+ year modernization effort.
Steal this: Pick one repetitive task your team completes manually every day (processing orders, updating records, checking inventory status). Use ChatGPT or similar AI to handle the routine cases for 2 weeks while your team reviews every output before it goes live. Track time saved weekly.
AI News Story
Scientists launch app that identifies dinosaur footprints
Researchers at the University of Edinburgh and Helmholtz-Zentrum have released DinoTracker, an AI-powered app that identifies dinosaur species from fossilized footprints with roughly 90% accuracy compared to human expert classification.
The system takes a different approach than previous AI tools. Rather than training on pre-labeled footprints (which may contain human errors), the team fed 2,000 unlabeled footprint silhouettes into the model. The AI then identified eight meaningful features on its own, including toe spread, ground contact area, and heel position.
The system has already added weight to a longstanding debate about birdlike tracks from the Triassic period, though researchers caution that human expertise remains necessary to verify factors like geological age and substrate material.
AI News Story
Google adds agentic vision to Gemini 3 Flash
Google has introduced Agentic Vision in Gemini 3 Flash, allowing the model to actively inspect images by planning steps, zooming into details, and running code to ground answers in visual evidence rather than static guesses.
By combining visual reasoning with code execution, the update delivers a reported 5–10% quality improvement across vision benchmarks, addressing long-standing issues like missed fine-grained details and visual hallucinations.
For developers, this signals a shift toward more reliable, audit-friendly computer vision workflows, with early adopters already using the feature for tasks like plan validation, image annotation, and visual math inside production systems.
Trending AI Tools (sponsored by our tools database)
A curated look at the AI tools quietly transforming how teams work.
Chatnode is for building custom, advanced AI chatbots that enhance customer support and user engagement ⭐️⭐️⭐️⭐️⭐️ (Product Hunt)
sshx is a collaborative web terminal with real-time sharing and end-to-end encryption
GiftAssitant curates personalized gift ideas based on recipient and occasion
Companies to Watch/Raising Now
Early-stage AI startups on our radar, before they’re everywhere.
• Ricursive Intelligence
What they do: Builds AI systems that co-design chips and models together, shortening hardware development cycles.
Why it matters: Raised $300M Series A just two months after launch, betting that AI-designed silicon will outpace human-led chip teams on speed and efficiency.
Stage: Series A
Investors: Lightspeed, Sequoia, Nvidia’s NVentures
• ZOHO.VC
What they do: Early-stage fund spun out of a German tech incubator, backing pre-Seed and Seed startups tied closely to operators.
Why it matters: Hit 70% of its €10M target at first close and already backed five startups, showing how incubator-native capital is moving faster than traditional funds.
Stage: Fund formation
Raising: €10M target
Know a company we should be watching? Hit reply and let us know.
Refer a Friend
We read your emails, comments, and poll replies daily.
Until next time, Martin, Arturo, and Liam.
P.S. Unsubscribe if you don’t want us in your inbox anymore.