AI Academy
Posts
🤖 ChatGPT Can Now Click Buy and Handle Your Tasks

🤖 ChatGPT Can Now Click Buy and Handle Your Tasks

Plus: Google Fights Back and Europe Splits the AI World

July 24, 2025

Hello AI Enthusiast,

This week shows where the AI industry is heading - and who's comfortable with that direction. OpenAI launched ChatGPT agent for web browsing and task automation, Google added phone-calling AI to search, and three major companies had very different reactions to Europe's new AI code of practice.

Let's break down what's happening and why it matters.

The Big Picture 🔊

OpenAI Launches ChatGPT Agent

OpenAI has released ChatGPT agent, which combines their previous Operator and deep research tools into one system that can browse websites, analyze data, and complete multi-step tasks like creating presentations or making purchases. Available now to Pro, Plus, and Team users in most regions (though not yet in the EU or Switzerland). OpenAI acknowledges the expanded capabilities introduce new risks and has implemented safety measures in response.

Andrea MattielloCM and Board Lover

OpenAI's showcasing some pretty ambitious examples - having AI handle financial modeling or book travel with real money involved. While impressive, these feel risky for something still in early days. We'd stick to safer tasks for now.

What's interesting is OpenAI making a play for enterprise users already locked into Microsoft or Google ecosystems. That bold claim about outperforming Copilot in Excel basically says "even if you're using Microsoft's tools, our AI works better than their built-in features."

Ready for the AI agent revolution? Our Corporate AI Training helps teams figure out where tools like ChatGPT agents actually add value versus where they just create more work. Most companies still struggle with basic AI - we teach you to spot the difference between impressive demos and practical solutions.

Google Adds Gemini 2.5 Pro and AI Phone Calling to Search

Google is bringing its smartest Gemini 2.5 Pro model and Deep Search capabilities to paid subscribers first, essentially creating a premium tier for search. Deep Search runs hundreds of searches to create comprehensive cited reports The more interesting feature: AI that calls local businesses for you. Search "pet groomers near me," hit "Have AI check pricing," and it calls multiple places to get you pricing and availability info. Rolling out to US users.

Helin YontarCPO and Polyglot

This feels like Google playing catch-up while trying to protect their business model. The phone calling feature is interesting, but it's a limited version of what OpenAI's agent declares it can do.

Google knows that if AI gets too good at answering questions directly, people stop clicking links and that kills their ad revenue. They're walking a tightrope between staying competitive and not destroying their core business.

Three AI Giants React Differently to EU's Code of Practice

The EU's AI code of practice is splitting tech companies. OpenAI and Anthropic committed to signing it, while Meta is flat-out refusing.

OpenAI's using this as a PR push, launching their "European Rollout" to build data centers and expand ChatGPT in schools. Anthropic says the code aligns with their safety principles. Meta called it an "overreach" that will "stunt" European AI development.

The code becomes binding next month, though signing remains voluntary for now.

James VarnhamCEO and Rainmaker

Meta's the odd one out, and it's not about AI safety - it's politics and business. The code isn't even legislation, just guidance for compliance with existing law.

The real issue? Meta makes money using AI to manipulate consumer behavior on their platforms. Any regulation limiting their ability to harvest European data threatens their core business model.

Bits and Bobs 🗞️

Mistral launched Voxtral, its first open-source audio model family, positioning it as an affordable alternative to closed corporate systems at less than half the price.
Runway launched Act-Two, a motion capture AI model that can animate any character from a single performance and translate it across different styles and environments.
Anthropic has launched a new directory of tools that connect to Claude, allowing users to integrate with services like Notion, Canva, and Stripe seamlessly.
OpenAI has enhanced its image generation API, allowing for higher fidelity editing of faces, logos, and specific details.
Netflix has used AI in the Argentinian series "El Eternauta" to enhance visual effects, achieving results 10 times faster and at a lower cost than traditional methods.
A group of AI researchers from leading AI companies are advocating for deeper investigation into "chain-of-thought" monitoring to enhance transparency and safety in AI reasoning models.

On the Podcast 🎧

In our latest episode of "I'm NOT a Robot", we chat with Ivo, an AI Academy graduate who moved to Costa Rica and built "Nosara is Calling" - a platform connecting tourists with local businesses. Using skills from our AI Agent Bootcamp (with zero coding experience), he created a solution that's helping his adopted hometown while letting him surf daily. Listen here.

That's a wrap on our newsletter! Before you go, here’s a quick recap of our offerings:

AI Academy Membership: Get 12 months of access to all our cohort-based programs, live webinars, on-demand courses, and tutorials.
AI Agent Bootcamp: Accelerate processes and solve business problems by mastering prompts and building AI Agents, without coding.
Practical Introduction to ChatGPT: A free course on using ChatGPT confidently, understanding its workings, and exploring its potential.
Customized Corporate Training: Equip your team with the skills they need to unlock the potential of AI in your business.

Catch you next week! 👋