Practical AI Newsletter12 August 2026
AI models shrink, agents grow up.

AI models shrink, agents grow up.

This week: AI models are becoming significantly more efficient and accessible, enabling powerful capabilities on smaller devices and expanding agentic applications. Liquid AI's LFM2.5-2.6B runs agents on hardware as small as a Raspberry Pi and Meta's Muse Glimmer ships as open weights you can run locally, while xAI's Grok 4.6 and Microsoft's MAI-Code-1.1-Flash both lift coding performance at the same or lower price. OpenAI also opened GPT-5.6 Luna to free users, so the no-cost option is no longer the weakest one.

Editor's Picks
Editor's Picks5

Introducing Grok 4.6

xAI released Grok 4.6, the successor to Grok 4.5, on August 12, 2026.

Expand details

MAI-Code-1.1-Flash cuts costs, speeds up coding

Microsoft announced on August 11, 2026, the release of MAI-Code-1.1-Flash, an updated coding model that delivers higher quality code, runs faster, and costs 75% less than its predecessor.

Expand details

Meta is back with Muse Glimmer: local, agentic, multimodal, and open source

Meta released Muse Glimmer, a new 30B parameter multimodal, local, agentic, and open-source model, on August 10, 2026.

Expand details

No cloud, no GPUs, no problem: Liquid AI's new model LFM2.5-2.6B brings powerful AI agents to devices as small as a Raspberry Pi

Liquid AI debuted LFM2.5-2.6B, a new open-weight language model designed for agentic workloads, earlier this week on August 6, 2026.

Expand details

Improving GPT‑5.6 Sol in ChatGPT—and expanding access to GPT-5.6 Luna for free users

OpenAI announced on August 6, 2026, updates to ChatGPT, improving GPT-5.6 Sol for Plus and Pro users and expanding access to GPT-5.6 Luna for free users.

Expand details
AI Business
AI Business8

Millennium and Anthropic are building a digital risk analyst with Claude

Anthropic announced on August 6, 2026, that it is co-developing a digital risk analyst with Millennium, one of the world's largest alternative investment management firms.

Expand details

In-region inference, open models, and new European infrastructure for sovereign AI.

Mistral AI announced on August 11, 2026, new initiatives for AI sovereignty, including generally available regional inference endpoints, expanded support for third-party open models, and a coalition to secure long-term European compute capacity.

Expand details

Microsoft expands AI model choice for startups

Microsoft for Startups announced resources for deploying Fireworks AI models on Microsoft Foundry, generally available on August 4, 2026.

Expand details

OpenAI’s letter to Governor Abbott on responsible AI infrastructure in Texas

OpenAI sent a letter to Texas Governor Greg Abbott outlining its commitment to responsible AI infrastructure development in Texas on August 10, 2026.

Expand details

The next chapter of our AI momentum

Google and Alphabet CEO Sundar Pichai announced organizational changes within Google DeepMind teams on August 5, 2026.

Expand details

Perspective

1

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

The Future is for Everyone

Mark Zuckerberg published a letter on August 10, 2026, outlining Meta's philosophy for a positive AI future focused on individual empowerment.

Expand details

AI & Society

2

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

From asking to doing: How the world is putting ChatGPT to work

OpenAI released new "Signals" data on August 6, 2026, detailing how people worldwide are using ChatGPT in their work and daily lives.

Expand details

Working with the American Psychological Association on youth mental health and AI

OpenAI announced on August 6, 2026, a partnership with the American Psychological Association (APA) to develop evidence-based guidance and safeguards for responsible AI use among young people.

Expand details
Product Updates
Product Updates22

AI Assistants

5

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

AI startup Hark unveils first product: an affordable, fast computer use agent Hark Handoff

AI startup Hark unveiled its first product, Hark Handoff, a computer use agent (CUA), on August 5, 2026.

Expand details

Ask Maps gets more helpful with food ordering and more

Google announced on August 6, 2026, that Ask Maps has been updated with new agentic capabilities, including food ordering, hotel discovery, and real-time transit information.

Expand details

Testing ads in ChatGPT

OpenAI announced on August 11, 2026, that it is expanding its pilot program for testing ads in ChatGPT to new international markets.

Expand details

More than 1 billion people are using the Gemini app every month.

Google announced on August 11, 2026, that the Gemini app has surpassed 1 billion monthly active users.

Expand details

Auto mode is now the default in Claude Code for Pro, Max, and Team plans

Anthropic made auto mode the default in Claude Code for Pro, Max, and Team plans, effective August 14, 2026.

Expand details

Productivity & Work

3

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

Introducing Grok Bot

xAI launched Grok Bot, an early beta of AI agents that work inside the apps a team already uses, on August 11, 2026.

Expand details

Computer for Builders

Perplexity announced "Computer for Builders," a collection of AI-powered tools for founders and small engineering teams, on August 6, 2026.

Expand details

Automatically add sources to your Gemini Notebooks in Workspace Studio

Google Workspace announced on August 7, 2026, a new integration in Workspace Studio that allows users to automatically add sources to their Gemini Notebooks.

Expand details

Creative Content

3

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

LTX-2.5 can generate a 10-second AI video from an image in just 6.8 seconds on Nvidia superchips — and it's open weights

LTX, an open world model company, released LTX-2.5, the newest version of its open-weights video and "world" model, on August 11, 2026, with native integration into ComfyUI.

Expand details

Introducing Adobe for ChatGPT: Create, edit and get work done — all in ChatGPT

Adobe introduced the Adobe plugin for ChatGPT, bringing its creative and productivity tools into ChatGPT, on August 6, 2026.

Expand details

Imagine Image 2.0

xAI released Imagine Image 2.0, a new quality mode for image generation and editing, on August 7, 2026.

Expand details

Marketing

1

Features for campaigns, content promotion, audience growth and reaching customers more effectively.

Evolve your marketing with new AI tools

Google announced new AI and agentic experiences across Google Ads and Google Analytics on August 10, 2026.

Expand details

Education & Learning

2

Features that help you learn new skills, study smarter and stay ahead with AI-powered training and courses.

Streamlining rubric generation in Google Classroom with Gemini

Google updated Google Classroom on August 5, 2026, to allow educators to generate rubrics with Gemini.

Expand details

Expanding the Google AI Professional Certificate with vibe coding

Google expanded its AI Professional Certificate on August 11, 2026, by introducing a new course dedicated to "vibe coding."

Expand details

Pricing & Access

2

What it costs, who can access it, and where it's available.

Copilot trial helps small businesses embrace AI

Microsoft rolled out "Copilot in 30," a 30-day trial of Microsoft 365 Copilot Business, on August 5, 2026.

Expand details

Premium seats are coming to ChatGPT Business

OpenAI announced Premium seats for ChatGPT Business, offering 5x more usage and removing the five-hour usage limit, on August 10, 2026.

Expand details

Other Updates

6

Noteworthy AI updates that don't fit neatly elsewhere.

Our WeatherNext 2 AI model demonstrated a massive leap forward in predicting cyclones.

Google DeepMind announced on August 6, 2026, that its WeatherNext 2 AI model achieved state-of-the-art accuracy in predicting tropical cyclones.

Expand details

AMIE, our research medical AI system, demonstrates real-time clinical video consultation capabilities in a first-of-its-kind study.

Google Research and Google DeepMind announced on August 11, 2026, advancements in AMIE, their research medical AI system, demonstrating expert-level real-time clinical video consultation capabilities in a simulated study.

Expand details

Expanding Daybreak as the Cyber Defense Window Narrows

OpenAI announced the expansion of Daybreak with two access tiers, Daybreak Blue and Daybreak Red, introducing GPT-5.6-Cyber for authorized vulnerability research, exploit validation, and security testing on August 10, 2026.

Expand details

Model ML completes finance work more efficiently with GPT-5.6 Sol

Model ML announced it uses GPT-5.6 Sol to complete finance work more efficiently, from research and analysis to editable PowerPoint decks and Excel workbooks, on August 10, 2026.

Expand details

Compliance API coverage extends to Claude Cowork and Claude Code

Claude announced on August 11, 2026, that its Compliance API now covers sessions in Claude Cowork across desktop, web, and mobile, and Claude Code in the CLI and desktop app, available in beta for Enterprise customers.

Expand details

Improving Fable 5's biology safeguards

Anthropic announced updates to Claude Fable 5's biology safeguards on August 7, 2026.

Expand details
AI in the World
AI in the World1

Microsoft’s newest India datacenter region goes live to power the country’s AI economy and enable Frontier Firms

Microsoft announced the general availability of its India South Central cloud region in Hyderabad, Telangana, on August 6, 2026.

Expand details
AI in Practice
AI in Practice13

Why AI-driven purchase intent so rarely becomes a completed sale

Rezolve Ai presented an article on August 11, 2026, discussing why AI-driven purchase intent often fails to convert into completed sales.

Expand details

TutorMoments: Do AI tutors know when to help and when to hold back?

Allen Institute for AI (AI2) introduced TutorMoments on August 7, 2026, a framework to measure whether large language models (LLMs) can effectively balance when to help a student and when to encourage independent reasoning in tutoring sessions.

Expand details

Case Studies

6

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

Virgin Atlantic sharpens customer journeys with ChatGPT Work

Virgin Atlantic announced on August 10, 2026, that it is using ChatGPT Work to accelerate research, product planning, and decision-making by connecting signals across the customer journey.

Expand details

How Zapier transformed core marketing processes with ChatGPT Work

OpenAI published a case study on August 10, 2026, detailing how Zapier's enterprise marketing team transformed core processes using ChatGPT Work.

Expand details

How HSP GRUPPE builds AI capabilities for tax advisory

OpenAI published a customer story on August 7, 2026, detailing how HSP GRUPPE, a network of tax advisory, auditing, and law firms, implemented ChatGPT Enterprise to enhance its AI capabilities.

Expand details

See what 5 builders are making with Gemini Omni

Google published an article on August 7, 2026, showcasing five different projects built by developers using Gemini Omni for video creation and editing.

Expand details

How two young Indonesians are building digital careers through AI and cloud

Microsoft published a story on August 4, 2026, detailing how two young Indonesians, Muhammad Yusuf Fadhel Marwiji and Emejleano Rusmin Nggepo, built digital careers using AI and cloud technologies through the Microsoft Elevate Training Center.

Expand details

Running auto mode in production

Claude Code launched auto mode as its default setting on August 7, 2026.

Expand details

Frameworks and Guides

5

How-to advice and frameworks you can apply yourself.

Your agent didn’t hallucinate; it exceeded its authority

VentureBeat published an article on August 10, 2026, highlighting that AI agent failures often stem from exceeding authority rather than hallucination, a governance gap many enterprises are not addressing.

Expand details

What building an AI-native finance function taught me

OpenAI CFO Sarah Friar shared five lessons for building an AI-native finance function on August 10, 2026.

Expand details

How to build an AI agent: A simple guide for anyone

Microsoft Signal Blog published a guide on how to build an AI agent using Microsoft 365 Copilot on August 10, 2026.

Expand details

How Gemini plans such detailed vacation itineraries for you

Google announced on August 6, 2026, how its Gemini AI assistant plans detailed and personalized vacation itineraries for users.

Expand details

Your AI agent may be ready. Your sales motion probably isn’t.

Salesforce presented an article on August 11, 2026, highlighting that many companies' sales motions are not prepared for the agent economy, even if their AI agents are ready.

Expand details
AI & Developer
AI & Developer3

Run Claude Code sessions on your own compute

Anthropic announced the public beta of self-hosted environments for Claude Code on August 6, 2026.

Expand details

Inference hooks: inline data loss prevention for Claude Enterprise

Anthropic launched Inference hooks for Claude Enterprise on August 5, 2026.

Expand details

Advance Zero Trust for AI: New tools and guidance to secure AI agents and DevSecOps

Microsoft announced new tools and guidance to advance Zero Trust for AI on August 4, 2026.

Expand details
Practical AI Newsletter

Join our free 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.