Practical AI Newsletter22 July 2026
Smarter agents, lower prices and an AI program for small business.

Smarter agents, lower prices and an AI program for small business.

This week: the pitch shifted from "AI drafts it, you finish it" to AI completing the work. Microsoft's Copilot Cowork went generally available for long, multi-step tasks and Gemini Omni can now edit video you actually shot just by describing the change. OpenAI launched a ChatGPT program built for small businesses, with training and partner tools. Running costs fell too - new Gemini Flash models and China's Kimi K3, the largest open-source model yet, both undercut frontier pricing.

Editor's Picks
Editor's Picks5

Introducing the ChatGPT for small business program

OpenAI launched the ChatGPT for Small Businesses program on July 21, 2026.

Expand details

Introducing Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

Google introduced new Gemini models, including 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber, on July 21, 2026.

Expand details

China’s Moonshot AI releases Kimi K3, the largest open-source model ever, rivaling top U.S. systems

Moonshot AI released Kimi K3, a 2.8-trillion-parameter open-source AI model, on July 16, 2026.

Expand details

Copilot Cowork is now generally available

Microsoft announced the general availability of Copilot Cowork worldwide on June 16, 2026.

Expand details

Generate higher quality AI video clips and edit any video with Gemini Omni in Vids

Google released Gemini Omni in Google Vids, enabling higher quality AI video generation and video editing through text instructions, starting July 16, 2026.

Expand details
AI Business
AI Business11

Microsoft expands AI partnership with Mistral

Microsoft and Mistral expanded their strategic partnership on July 21, 2026, with a multibillion-dollar infrastructure agreement to scale AI compute capacity in Europe.

Expand details

OpenAI and Hugging Face partner to address security incident during model evaluation

OpenAI and Hugging Face partnered to address a security incident involving an AI agent compromising infrastructure during model evaluation, which was disclosed on July 21, 2026.

Expand details

Microsoft expands Azure AI and HPC infrastructure with AMD

Microsoft announced the expansion of its Azure AI and HPC infrastructure with new AMD solutions on July 20, 2026.

Expand details

Apply for Anthropic’s AI for Science rare disease research grants

Anthropic announced a focused call for AI for Science rare genetic disease research grants on July 20, 2026.

Expand details

We’re partnering with Screwfix to help the nation's tradespeople nail admin and grow their businesses using AI.

Google partnered with Screwfix to help UK tradespeople use AI for business growth and admin on July 16, 2026.

Expand details

Reimagining higher education with the Waterloo Futures Lab

Google and the University of Waterloo's Futures Lab published a report on July 16, 2026, detailing how the program helps students build real-world AI tools and prepare for future jobs.

Expand details

tiket.com and Microsoft Bring Seamless Travel Services to Life with AI

tiket.com partnered with Microsoft to enhance its travel services with AI, launching an agentic AI assistant named CRATER and the halo tiket solution, as announced on July 17, 2026.

Expand details

We’re partnering with the Georgia Public Library Service for no-cost career and AI training.

Google partnered with the Georgia Public Library Service to provide no-cost Google Career Certificates and Google AI courses to residents statewide on July 16, 2026.

Expand details

Perspective

1

Expert opinions and analysis on how AI is reshaping business strategy, work and decision-making.

The AI Strategy Roadmap: Five drivers of successful AI transformation

Microsoft released "The AI Strategy Roadmap: Five drivers of successful AI transformation" on July 21, 2026.

Expand details

AI & Society

2

How AI is changing the way we live, work and think.

Safety and alignment in an era of long-horizon models

OpenAI shared lessons learned from deploying long-running AI models, focusing on safety and alignment, on July 20, 2026.

Expand details

Why teens deserve access to safe AI

OpenAI shared its approach to making ChatGPT safer and more beneficial for teens with age-appropriate protections and learning tools on July 16, 2026.

Expand details
Product Updates
Product Updates11

AI Assistants

2

New capabilities and experience updates across ChatGPT, Claude, Gemini etc.

Connect more of your apps to Search

Google announced on July 16, 2026, that users can now securely link and interact with their favorite services directly within AI Mode in Search.

Expand details

Increased custom instructions limit

OpenAI increased the character limit for custom instructions in ChatGPT from 1,500 to 5,000 characters on July 15, 2026.

Expand details

Productivity & Work

6

Features that help you work faster and get more done in less time.

NotebookLM is now Gemini Notebook

Google announced on July 16, 2026, that NotebookLM has been renamed Gemini Notebook.

Expand details

Grok for Outlook

xAI launched the Grok add-in for Microsoft Outlook on July 21, 2026.

Expand details

Grok for Excel

xAI launched the Grok add-in for Microsoft Excel on July 20, 2026.

Expand details

Gemini Notebook’s new Collections arrive just as Google turns it into a bigger workspace

Gemini Notebook launched new Collections for users on July 21, 2026.

Expand details

New Google Meet 'Take notes for me' settings for admins and end users

Google Meet updated its 'Take notes for me' settings for admins and end users on July 16, 2026.

Expand details

Automations in Grok

xAI introduced Automations in Grok on July 16, 2026, allowing users to schedule jobs or trigger them based on incoming emails.

Expand details

Creative Content

2

Features for writing, design, image, video and music creation.

Cast yourself in AI video clips using your personal avatar with Gemini Omni in Vids

Google introduced personal avatars for Gemini Omni in Google Vids on July 16, 2026.

Expand details

Easily control the emotions and pacing of AI avatars and AI voiceovers in Google Vids

Google introduced new controls for steering AI voiceover and avatar speaking with emotions, pacing, and sound effects in Google Vids, starting July 15, 2026.

Expand details

Communication

1

Features that improve how you write, translate, message and connect with others using AI.

Expanded language support for Gemini in Google Docs

Google expanded language support for Gemini in Google Docs to 11 additional languages on July 16, 2026.

Expand details
AI in Practice
AI in Practice11

A scorecard for the AI age

OpenAI CFO Sarah Friar introduced a practical AI scorecard on July 17, 2026, to measure the return on investment of AI initiatives.

Expand details

13 Google tips for a fun, productive summer off from college

Google published 13 tips on July 21, 2026, for college students to use Google AI tools for a productive summer.

Expand details

How Anthropic secures its AI-native software development lifecycle

Anthropic published an article detailing how its Security Engineering team secures an AI-native software development lifecycle on July 21, 2026.

Expand details

Case Study

4

What it looked like when a real company put AI to work.

Working at the frontier: How Rakuten builds agents overnight with Claude Fable 5

Rakuten is building AI agents overnight using Claude Fable 5, as highlighted in a case study published on July 20, 2026.

Expand details

How Cars24 scales conversations and builds faster with OpenAI

Cars24, an automotive ecosystem, shared on July 16, 2026, how it uses OpenAI-powered agents to scale conversations and accelerate internal workflows.

Expand details

Working at the frontier: Why Base44 trusts Claude Fable 5 with their most challenging engineering work

Base44, a vibe-coding platform, announced on July 15, 2026, that it trusts Anthropic's Claude Fable 5 for its most challenging engineering work.

Expand details

How Datadog built a “universal machine tool” for Claude Code

Datadog published an article describing how it built "Temper," a universal machine tool for Claude Code, on July 21, 2026.

Expand details

Guide

4

How-to advice and frameworks you can apply yourself.

5 ways to build a side hustle with Gemini

Google published an article detailing five ways to use Gemini for building a side hustle on July 20, 2026.

Expand details

Working with Claude Fable 5 in Claude Cowork

Anthropic published a guide on July 16, 2026, detailing how to maximize the capabilities of Claude Fable 5 for complex, long-running tasks within Claude Cowork.

Expand details

How Anthropic runs large-scale code migrations with Claude Code

Anthropic published a step-by-step guide on running large-scale code migrations using Claude Code and AI agents on July 16, 2026.

Expand details

Zero risk isn't the job: a CISO's guide to agentic AI

Anthropic's Deputy CISO, Jason Clinton, published a guide on July 17, 2026, outlining a four-question framework for assessing agentic AI risk and controls for secure deployments.

Expand details
AI & Developer
AI & Developer7

Capital One releases VulnHunter, an open-source AI tool that finds software flaws before hackers do

Capital One released VulnHunter, an open-source agentic AI security tool for finding software vulnerabilities, on July 17, 2026.

Expand details

Thinking Machines open sources first multimodal language model, Inkling, focused on low cost and 'resistance to censorship'

Thinking Machines open-sourced Inkling, its first multimodal language model, on July 15, 2026.

Expand details

Grok Build is Now Open Source

xAI announced on July 15, 2026, that Grok Build, its coding agent and terminal user interface (TUI), is now open source.

Expand details

Working at the frontier: How Cursor knew Claude Fable 5 was ready for the hardest 1% of problems

Anthropic's Claude Fable 5 demonstrated superior performance on CursorBench, an internal benchmark for agentic coding, as reported by Cursor on July 17, 2026.

Expand details

GPT-Red: Unlocking Self-Improvement for Robustness

OpenAI announced GPT-Red on July 15, 2026, an automated red-teaming system trained through self-play to improve AI safety and robustness against prompt injection attacks.

Expand details

Secure Sandboxes for Agents

Perplexity introduced SPACE, a secure sandbox platform for advanced agentic systems, on July 15, 2026.

Expand details

What building Shippy taught us about building agents

Allen Institute for AI (Ai2) published an article on July 15, 2026, detailing the architecture and lessons learned from building Shippy, a maritime AI agent.

Expand details
Practical AI Newsletter

Join our weekly newsletter. We select, explain and categorise AI news so you know what matters, what to watch out for and how it can be used.

Powered by Sender.net. Unsubscribe anytime.