AI News

Apple Music to Label AI-Generated Songs Globally — "Made With AI" Tag Visible to All Users

                                          

Daily Tech Reader - AI Edition


Podcast 🎧 • Video 📽 • Short 📽


Apple

Apple

Apple Music to Label AI-Generated Songs Globally — "Made With AI" Tag Visible to All Users

Apple notified music industry partners today that it will apply a visible "Made With AI" label to content materially generated by AI, building on Transparency Tags announced in March that let creators disclose AI use in artwork, lyrics, and music video.

Sources: MacRumors · The Hollywood Reporter

Siri AI Goes Live Across All Apple Platforms in iOS 27 Beta — Powered by Apple Foundation Models Built on Gemini

The rebuilt Siri AI is now available for developer testing across iOS 27, iPadOS 27, macOS Golden Gate, and visionOS 27, with Apple's Foundation Models trained in collaboration with Google running on-device and via private cloud.

Sources: Apple · 9to5Mac · TechCrunch

Apple in Talks With News Publishers on Real-Time Siri AI Licensing Deals — Flashback to 2024 Headline Debacle

The Wall Street Journal reports Apple is negotiating licensing deals with news organizations to feed live information into Siri AI, reviving the strategy that collapsed when Apple Intelligence falsely attributed a BBC headline in 2024.

Sources: Wall Street Journal · Gizmodo


Microsoft

Microsoft Copilot Reaches 30 Million Paid Seats — Azure Up 43% as OpenAI Becomes 365 Subprocessor

Microsoft closed FY26 with 30 million Copilot seats and net adds more than doubling quarter-over-quarter, while OpenAI was quietly added as a Microsoft 365 subprocessor with the toggle auto-enabled for most tenants.

Sources: Empowering Cloud · GeekWire

Microsoft Launches Guild Copilots for Healthcare and Legal — Industry-Specific AI Tier Expands After Successful Pilots

After successful pilots in healthcare and legal, Microsoft formally announced a new industry-specific Copilot tier in August, with Guild Copilots providing vertically tuned AI embedded directly in sector workflows.

Sources: AI Conference London · Microsoft

Copilot in SharePoint Now Generates Live HTML Dashboards Directly From Lists — No Rebuild Required

Copilot can now generate interactive HTML dashboards from SharePoint lists that stay connected to the underlying data, automatically refreshing whenever the list updates without requiring a manual export-and-rebuild cycle.

Sources: Microsoft Community Hub


Claude

Claude Autonomously Designs Protein Binders for 14 of 15 Targets — Hits Double the Industry Rate in Wet-Lab Validation

Anthropic's Claude Science ran an autonomous protein design campaign generating 1,320 candidates across 15 biological targets, with 354 confirmed binders at hit rates of 22–35% — more than doubling the 10–15% industry baseline in independent tests by Adaptyv Bio and Twist Bioscience.

Sources: Anthropic · Pharmaphorum · The Next Web

Anthropic Clarifies Watermark Scope After User Confusion — Even Proofreading a Paragraph Can Leave a Trace

Following user confusion over the August 2 global watermarking rollout, Anthropic clarified that the invisible mark indicates Claude had any role in the output — including light editing or translation — not that it generated the entire piece.

Sources: Fortune · The Hindu

Dual-Use Concern Flags Autonomous Bio-AI — Anthropic Blocks Protein Design on Fable 5, Builds Vetted Access Program

Anthropic acknowledged the same protein-design capabilities that accelerate medicine could enable bioweapon development, keeping those features locked on Claude Fable 5 and restricting access to a vetted scientist program while it builds safer release pathways.

Sources: Anthropic · The Next Web · TechTimes


ChatGPT

GPT-5.6 Luna Becomes ChatGPT Free Default — Two Cheaper API Tiers Cut Up to 80% on July 30

OpenAI named GPT-5.6 Luna the default model for free-tier ChatGPT users and removed the text-chat cap, after slashing Luna API pricing 80% and Terra 20% on July 30 while keeping the flagship Sol tier unchanged at $5/$30 per million tokens.

Sources: AI Weekly · FelloAI

OpenAI's "Sign in With ChatGPT" Expands to Airtable, GitLab, HubSpot, Notion, and More

OpenAI's identity layer beta, launched July 29, is expanding to additional partners, positioning ChatGPT as a front-door login for third-party tools by sharing only name, email, and profile picture.

Sources: ReconnAI · AI Weekly

OpenAI Math Paper Claims Ten Advances in Mathematics and Theoretical Computer Science

OpenAI published a research paper on August 1 claiming ten distinct advances in mathematics and theoretical computer science, continuing its push into formal reasoning benchmarks alongside the GPT-5.6 launch.

Sources: OpenAI · AI Weekly


Gemini

Gemini Hits 1 Billion Monthly Active Users — Google Reshuffles DeepMind Leadership to Close Gap With OpenAI

Google's Gemini app crossed 1 billion monthly users as of mid-August, while Alphabet restructured its AI leadership by centralizing Google DeepMind under new chief Koray Kavukcuoglu, who inherits the task of closing the frontier model gap with OpenAI and Anthropic.

Sources: CNBC · AIToolsRecap · Bloomberg

Google Renames NotebookLM to Gemini Notebook — Adds Secure Cloud Computer for Native Code Execution

NotebookLM is now Gemini Notebook, gaining a secure cloud computer inside every notebook for native code execution and data analysis, with cross-app syncing to the Gemini app rolling out to AI Ultra and Workspace Business users today.

Sources: AI Weekly · Google


China AI

DeepSeek Raises API Prices 93% After Race-to-Zero Backlash — V4 Flash Now at $0.27 Per Million Tokens

DeepSeek raised its V4 Flash API pricing 93% to $0.27/M tokens on August 17, reversing the aggressive undercutting strategy that had pressured U.S. labs into slashing mid-tier rates and accelerated what analysts called an "AI race to zero."

Sources: AIToolsRecap · Axios

Google DeepMind and Isomorphic Launch Bioresilience Program — AI Foundation Models Aimed at Pathogen Surveillance and Vaccine Design

DeepMind and Isomorphic Labs launched a Bioresilience initiative applying AI foundation models to pathogen surveillance and vaccine design, positioned as a counterweight to dual-use biological AI risks as Hassabis warned frontier bio-AI threats could reach open-source models within 18 months.

Sources: AI Weekly · Google DeepMind


Big Story

The Protein Design Moment: Claude's Autonomous Biology Results Signal a New Phase for AI in Drug Discovery

Anthropic's Claude Science hitting 14 of 15 protein binder targets at double industry rates — validated independently in a wet lab — marks a measurable turning point for autonomous AI in life sciences, even as biosafety concerns force restricted access to the most capable models.

Sources: Anthropic · Pharmaphorum · Dataconomy

EU DMA Orders: Claude, ChatGPT, and Copilot Must Get System-Level Android Access by August 2027

New EU Digital Markets Act binding orders require Google to grant Claude, ChatGPT, and Copilot system-level Android access by August 2027, while also mandating Google share search data with rivals starting January 2027.

Sources: AIToolsRecap · EU Commission



Copyright © 2026 Daily Tech Reader
All Rights Reserved

Popular posts from this blog