Latest AI Productivity News

Curated daily - the AI tools, productivity launches, automation updates, and platform releases worth your time. We link straight to the original source.

DevelopersIO
Anthropic Claude

DevelopersIO

Claude Code v2.1.232 Makes Subagent Forking the Default, Adds Cross-Session @Mentions, and Caps Runaway Search and Delegation Loops

The August 14 Claude Code release sets subagent forking on by default so a fork-type subagent inherits the full conversation and prompt cache without extra configuration, and introduces cross-session @mentions — type @ in the prompt to target another Claude session by name and Claude will route the message via SendMessage. The update also adds per-session caps of 200 on both WebSearch calls and subagent spawns to prevent runaway loops, with both limits tunable via environment variables.

Aug 14, 2026

MacRumors
Business AI

MacRumors

Apple Trains Its Own AI Model for the Chinese Market With Alibaba Support, Becoming the First Foreign Company Approved to Offer a Proprietary LLM in China

Apple has developed a proprietary large language model for China with help from Alibaba, breaking from its previous approach of licensing a domestic model to power Apple Intelligence in the country. The move makes Apple the first foreign company approved by the Chinese government to offer its own AI model there, with deployment expected alongside a fall iOS update coinciding with new iPhone launches.

Aug 14, 2026

Caixin Global
AI Tools

Caixin Global

DeepSeek Releases V4-Pro With 1M-Token Context and Agent Focus — Then Raises API Prices by Up to 1,100%

DeepSeek officially launched V4-Pro on August 14 — a 1-million-token context model built for agentic workflows including tool use, code execution, and multi-step reasoning — simultaneously announcing steep API price increases, with output tokens rising from $0.87 to $3.96 per million at peak hours, a more than 1,100% hike. DeepSeek also introduced peak and off-peak billing, with off-peak rates at half the peak-hour price.

Aug 14, 2026

TechCrunch
Business AI

TechCrunch

IBM Partners With OpenAI to Embed GPT-5.6 and ChatGPT Work Across IBM Consulting Advantage, Training Tens of Thousands of Consultants

IBM announced a strategic partnership with OpenAI to integrate GPT-5.6, Codex, and ChatGPT Work directly into IBM Consulting Advantage — its enterprise delivery platform serving financial services, government, telecom, and retail — and will create a dedicated OpenAI Practice to train and certify tens of thousands of IBM consultants on OpenAI technologies. The deal positions IBM as a major enterprise distribution channel for OpenAI as both companies push deeper into AI-driven professional services.

Aug 13, 2026

Super Simple 365
Microsoft Copilot

Super Simple 365

Microsoft Ships Eight Copilot Updates in August, Adding Notebooks for All Users and a New Workflow Automation Tab

Microsoft delivered eight Microsoft 365 Copilot updates this month, including Copilot Notebooks for all subscribers, faster Copilot connector content crawl, inline rich image rendering in chat, and a new Cowork integration tab for building AI-powered workflow automations from internal organizational data. The release marks a shift in Copilot's identity from a text assistant toward a cross-application automation hub.

Aug 12, 2026

TechCrunch
Anthropic Claude

TechCrunch

Anthropic Begins Watermarking All Claude-Generated Text With Invisible Machine-Readable Marks Under EU AI Act Rules

Anthropic announced on August 11 that every Claude model launched on or after August 2, 2026 now embeds an imperceptible statistical watermark directly into generated text — covering the API, claude.ai, Claude Code, Claude Cowork, Claude Tag, and Claude accessed via AWS, Google Cloud, and Microsoft Foundry. The mark travels with copied and pasted text in response to Article 50 of the EU AI Act's transparency requirements, though Anthropic notes it cannot confirm who was responsible for the output or whether it was subsequently edited.

Aug 11, 2026

Anthropic
Anthropic Claude

Anthropic

Unreleased Claude Model Raises Riemann Hypothesis Zeta Bound from 41.6% to 67.2%, the Largest Single Advance in the Problem's History

An experimental pre-release Claude model connected findings across two mathematics sub-fields to raise the proven lower bound on the fraction of nontrivial Riemann zeta zeros satisfying the hypothesis from 41.6% to 67.2% — the largest single-step gain on this measure ever recorded — achieved in 36 hours using a swarm of autonomous sub-agents. Anthropic mathematicians verified the result by hand, outside experts reviewed the approach, and Claude generated a machine-checkable Lean proof; Anthropic noted the technique is not expected to prove the full Riemann hypothesis.

Aug 11, 2026

TechCrunch
OpenAI / ChatGPT

TechCrunch

OpenAI Expands Daybreak Cybersecurity Program and Launches GPT-5.6-Cyber, a Model Purpose-Built for Authorized Security Work

OpenAI expanded its Daybreak cyber defense initiative on August 10 by launching GPT-5.6-Cyber — a purpose-trained model that responded to 95% of requests tied to advanced security work including exploit-chain development and authentication bypass — structured into two access tiers (Daybreak Blue for general defensive work, Daybreak Red for authorized vulnerability research) with new partners including Accenture, IBM, CrowdStrike, Cisco, Sophos, and Cloudflare. The models became available on Amazon Bedrock on August 11, and all individual Daybreak accounts must adopt hardware security keys by September 1.

Aug 10, 2026

Meta AI Research
AI Tools

Meta AI Research

Meta Releases Muse Glimmer, a 30B Open-Weight Agentic AI Model That Runs on a Single Consumer GPU

Meta Superintelligence Labs released Muse Glimmer on August 10 — a 30-billion-parameter open-weights model (Apache 2.0) built for local, always-on agent workflows that supports long-horizon reasoning, reliable tool calling, failure recovery, and multimodal reasoning across more than 100 languages. The model runs on a single consumer GPU, is available on Hugging Face, and CEO Mark Zuckerberg used the launch to call on the US government to reduce barriers to open-source AI development.

Aug 10, 2026

Bloomberg
Business AI

Bloomberg

Anthropic, Macquarie, and GIC Launch Theseus Infrastructure to Build Dedicated AI Data Centers Across the United States

Anthropic, Macquarie Asset Management, and Singapore's GIC announced Theseus Infrastructure on August 10 — a joint venture to develop, operate, and lease US data center campuses purpose-built for Claude's compute needs, with Macquarie and GIC funding the majority of equity while Anthropic serves as anchor tenant. Anthropic committed to covering 100% of grid-upgrade costs and any consumer electricity price increases linked to its demand, extending its infrastructure build-out alongside the $10 billion Volta Infra deal signed the previous week.

Aug 10, 2026

OpenAI
OpenAI / ChatGPT

OpenAI

OpenAI Launches $125/Month Premium Seats for ChatGPT Business, Offering 5x More Usage and No Five-Hour Throttle

OpenAI introduced Premium seats for ChatGPT Business at $125 per user per month (or $100 billed annually), giving team members five times the usage allowance of Standard seats and removing the five-hour usage cap that throttled heavy business users — while Standard seats remain at $25/month. Workspaces that add Premium seats before August 20, 2026 can receive $100 in workspace credits per qualifying seat, up to five seats.

Aug 10, 2026

Anthropic
Business AI

Anthropic

Anthropic Opens Claude for Government Desktop in Public Beta with FedRAMP High Auth, Bringing Claude Code and Cowork to Federal Agencies

Anthropic launched Claude for Government Desktop in public beta, making Claude Code and Claude Cowork available inside a FedRAMP High-authorized environment — with tamper-evident audit logs, admin spend controls, and department-level configuration for agency IT teams. Federal agencies can request access at claude.com/solutions/government, with a limited-time program offering the platform at $1 per agency with unlimited seats through August 2026.

Aug 10, 2026

Google Workspace Updates
Google Gemini

Google Workspace Updates

Google Rolls Out Gemini in Classroom to Students of All Ages, Adding AI-Powered Flashcards and Practice Quizzes from Class Materials

Google began rolling out Gemini in Classroom on August 10 to K-12 and higher education students of all ages, giving them a dedicated Gemini tab where they can transform course materials into AI-generated flashcards and practice quizzes — without student data being used to train AI models. Schools retain full control through admin settings, and mobile access follows on August 17.

Aug 10, 2026

TechCrunch
Anthropic Claude

TechCrunch

Anthropic Sets Auto Mode as Default for Claude Code on August 14, After Study Shows AI Catches 89% of Harmful Actions vs. 13.6% for Humans

Starting August 14, Claude Code on Pro, Max, and Team plans will default to auto mode — letting the AI execute tasks without human approval except for actions it classifies as irreversible or destructive. Anthropic justified the shift with internal research finding its auto-mode classifier caught 89% of harmful actions while human review caught only 13.6%, suggesting supervised autonomy is safer than manual approval workflows in practice.

Aug 9, 2026

OpenAI
OpenAI / ChatGPT

OpenAI

OpenAI Will Retire the Official DALL-E GPT from ChatGPT on August 30, Consolidating Image Generation Under ChatGPT Images

OpenAI announced it will remove the official DALL-E GPT from ChatGPT on August 30 and end DALL-E model access in Codex on August 31, directing all image creation to ChatGPT Images, which now serves as the primary hub for generating and editing AI images inside the platform. User-created GPTs with image generation are unaffected, but anyone with saved DALL-E GPT images should download them before the deadline.

Aug 9, 2026

TechCrunch
OpenAI / ChatGPT

TechCrunch

OpenAI Pauses Astra Development After Model Hits 'Critical' Cybersecurity Capability Threshold

OpenAI suspended internal activities tied to its upcoming Astra model on August 7 after an internal review found it could independently identify and execute zero-day exploits against hardened real-world systems — meeting the 'Critical' cybersecurity threshold in OpenAI's Preparedness Framework for the first time by any frontier lab. The company is expanding isolated testing environments and will not release Astra until stricter security controls are fully in place.

Aug 7, 2026

TechCrunch
AI Tools

TechCrunch

Cloudflare Launches Kitesurf, a Free Browser Built for AI Agents to Navigate and Automate Web Tasks

Cloudflare launched Kitesurf in free beta on August 7, a headless browser purpose-built for AI agents to navigate websites, fill out forms, and complete browser-based tasks — available via Browser Run in its Developer Platform. Unlike conventional headless browsers, Kitesurf is designed specifically for agentic use cases, letting developers build AI-powered software that acts on the open web without requiring standalone infrastructure.

Aug 7, 2026

One AI workflow, every weekday.

Tutorials, tool reviews, and automation playbooks for solopreneurs running on AI. Short, useful, and free. Unsubscribe anytime.

No pitch. No upsell. One quick AI workflow per weekday.