Hello AI Enthusiast,
This week, OpenAI, Microsoft, Anthropic, and Google all made major announcements. The common thread? Getting deeper into the tools you use at work, from spreadsheets to code reviews to full task execution. The names change, the pitch stays the same. Let's break down what actually matters.
The Big Picture 🔊
OpenAI Launches GPT-5.4 and Brings ChatGPT Into Excel
OpenAI has released GPT-5.4, a reasoning model that combines coding, computer use, and document work into one. The company claims improvements in accuracy, web research, and tool efficiency compared to previous versions. Alongside it, OpenAI launched a beta Excel add-in powered by GPT-5.4 that lets users build and update spreadsheet models through chat. New integrations with financial data providers like Moody's and Dow Jones Factiva are also included.
|
||
|
A few weeks ago, Claude launched its own Excel integration. Now OpenAI is doing the same. The pattern is clear: every major AI company is racing to embed itself into the tools people already use at work, and right now, spreadsheets are the main battleground. |
Anthropic Launches Code Review for Claude Code and a Partner Marketplace
Anthropic has introduced two new enterprise features. Claude Code now includes an AI-powered code review system that dispatches a team of agents on every pull request to find and rank bugs, costing $15-25 per review. Separately, Anthropic launched the Claude Marketplace, letting enterprise customers use their existing Anthropic commitment to pay for Claude-powered partner tools from companies like Harvey, GitLab, Snowflake, and Lovable, without separate procurement processes.
|
||
|
The marketplace idea makes sense for enterprise buyers who already have an Anthropic spending commitment and want to simplify procurement. The code review feature is more debatable. Similar tools already exist at a fraction of the cost, and at $15 to $25 per review, the pricing adds up fast. Claude's tendency to overthink and spin up multiple agents before completing a task does not help the case here. |
Microsoft Copilot Gets an Execution Mode With Cowork
Microsoft has introduced Copilot Cowork, a feature that moves beyond answering questions to actually completing tasks inside Microsoft 365. Users describe what they need, and Cowork creates a plan, then executes it across Outlook, Teams, Excel, and other apps, with user approval at each step. Use cases include calendar management, meeting prep, company research, and product launches. Built in collaboration with Anthropic, it is currently in limited Research Preview.
|
||
|
Copilot Cowork moves AI from answering questions to actually completing tasks across your Microsoft 365 apps. You describe the outcome, it builds a plan and executes it, checking in before making changes. The real test is whether people will trust it enough to genuinely delegate work, rather than just chat. |
Every major AI company is now embedded in your work tools. OpenAI is in Excel, Google is in Docs and Sheets, Microsoft is automating your workflows. Your team is already using AI whether you planned for it or not. Our Corporate AI Training helps companies move from accidental usage to a clear, practical strategy.
Bits and Bobs 🗞️
Google has updated Gemini across Docs, Sheets, Slides, and Drive, moving it from a sidebar assistant to a more integrated content partner that pulls context from your emails, files, and chats to help you create and edit directly inside each app.
OpenAI's new Codex Security uses AI with deep project context to identify and fix complex software vulnerabilities while reducing false positives and speeding up secure code delivery.
Anthropic's Claude Opus 4.6 independently found over 100 novel high-severity software vulnerabilities in Firefox, demonstrating AI's growing ability to identify security flaws faster than exploit development.
Anthropic is suing the U.S. government to block its designation as a security risk that restricts the use of its AI by federal agencies, arguing it violates constitutional rights and could harm its business.
Meta acquired Moltbook, a platform connecting AI agents through an always-on directory, to advance innovative and secure AI experiences.
Zoom is launching AI-powered avatars, office apps, and an AI agent builder for meetings, along with deepfake detection and AI-driven productivity tools across its platform.
YouTube is piloting AI deepfake detection technology for government officials, political candidates, and journalists to help identify and manage AI-generated impersonations while balancing free expression.
LOLgorithms 😂
Copilot Cowork is great, but… can you fix this first?
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 and on-demand courses.
AI Agent Bootcamp: Accelerate processes and solve business problems by 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! 👋
