AI News
Daily AI industry news — funding, products, policy, and major moves from global AI companies and startups.
SpaceX AI Device: The Quest for a Post-iPhone Handset
SpaceX is reportedly developing an AI handset thinner than the iPhone. The device aims for vertical integration using xAI and a custom OS. M
Benjamin Hollon Launches Writing Month to Combat AI-Induced Block
Benjamin Hollon launched Writing Month to support creators facing AI pressure. The initiative counters the devaluation of human creativity b
Meta Deploys AI Gateway to Curb 73.7 Trillion Token Waste
Meta introduced an AI Gateway to monitor massive token consumption. Employees used 73.7 trillion tokens in 30 days, prompting cost controls.
The Control Gap: Why 85% of Enterprises Struggle with AI Governance
85% of enterprises operate multiple competing AI platforms simultaneously. A critical control gap exists between AI adoption speed and gover
Claude Fable 5 Returns With a 99% Jailbreak Block Rate
Anthropic redeployed Claude Fable 5 globally after US export controls ended. A new safety classifier blocks over 99% of known jailbreak tech
Flint: The Qwen 3-Based Model Breaking the Average Answer Trap
Springboards developed Flint based on the Qwen 3 open-source model. The model uses targeted randomness to avoid predictable AI responses. Cr
Anthropic's Claude Fable 5 and Mythos 5 are finally breaking free
The US government has lifted export restrictions on Claude Fable 5 and Mythos 5. Fable 5 is now available globally, while Mythos 5 access is
Square Integrates ChatGPT and Claude to Replace 30% Delivery Fees
Square launched ChatGPT and Claude integrations for food merchants. The system replaces 30% delivery fees with a 2.9% transaction cost. Real
NS Hub Brings Claude Code Control to Mobile Browsers
NS Hub allows developers to control Claude Code via mobile browsers. The system uses Stone and ConceptGraph to maintain task context. The pr
Claude Fable 5 Restores Global Access After US Export Control Lift
Anthropic restored global access to Claude Fable 5 after US export controls were lifted. The model enabled Stripe to migrate 50 million line
Gemini Spark for macOS Turns Local Files Into an AI-Driven Workspace
Google launched Gemini Spark for macOS to automate local file management. The agent integrates with third-party apps and supports Model Cont
Ornith-1.0 Brings Autonomous Coding Local With MIT License
Ornith-1.0 is an autonomous coding model available under the MIT license. The model outperforms Qwen and Gemma on terminal and repository be
Show GN Maps the Branching Conversation Paths of Claude and ChatGPT
Show GN is a Chrome extension that visualizes AI chat branches as tree graphs. The tool supports both Claude and ChatGPT through a custom da
Anthropic Launches Claude Science to Supercharge Research
Anthropic introduced Claude Science, a specialized environment for scientific data analysis. It aims to reduce hallucinations and automate t
Why Morgan Stanley Limited FIXR's Autonomy to Save 1,500 Hours Weekly
Morgan Stanley deployed the FIXR AI agent for P&L reconciliation. The system reduces ledger processing time from six hours to three. Human f
Godot Foundation Bans AI-Generated Code to Combat Reviewer Fatigue
The Godot Foundation has banned AI-generated code contributions. The policy aims to reduce reviewer fatigue caused by AI slop. Human account
Etched Secures $1 Billion in Orders to Challenge Nvidia's Inference Dominance
Etched has secured $1 billion in orders for its specialized inference clusters. The startup reached a $5 billion valuation following a $500
GEO Strategy Pushes Dental Clinic AI Recommendation Rate to 27%
Generative Engine Optimization increased a dental clinic's AI visibility to 27%. The study tested ChatGPT, Claude, Gemini, and Perplexity us
Why the US Government Just Unlocked Anthropic's Mythos and Fable
The US government removed export license requirements for Anthropic's Mythos and Fable. Global access to these high-performance models resum
Wayve Leverages $8.5 Billion Valuation for $85 Million Talent Retention
Wayve is conducting an $85 million tender offer for its employees. The company is valued at $8.5 billion following a massive Series D round.
Why Claude Code 2.1.196 Hides Environment Signals in System Prompts
Claude Code 2.1.196 embeds hidden signals in its system prompts. The tool fingerprints users who override the official API URL. Hidden Unico
Tidal Cuts Royalties for AI-Generated Music to Protect Human Creators
Tidal is ending royalty payments for AI-generated music. The platform will introduce AI identification tags in mid-July. Distributors must n
Pi Coding Agent Scales with New pi-subagents Delegation Tool
Pi released pi-subagents to allow coding agents to delegate tasks. The tool uses git worktree to prevent file conflicts during parallel exec
Atlassian's DESIGN.md Experiment to Combat Generic AI UI Slop
Atlassian introduced DESIGN.md to prevent generic AI-generated interfaces. The format combines design tokens with rationale to preserve bran
OpenClaw Brings Open-Source AI Agent Control to iOS and Android
OpenClaw launched open-source AI agent apps for iOS and Android. The system uses a Gateway routing layer to connect mobile devices to tools.
The AI Development Harness That Let One Engineer Build a Monorepo
A custom AI harness enabled a team to scale deployments from 6 to 400 per week. One engineer integrated 95% of a monorepo in one month using
Claude Science: Why Anthropic Is Betting on Workflows Over Model Benchmarks
Anthropic launched Claude Science as an integrated workbench for researchers. The platform integrates over 60 scientific databases using a m
Claude Science Integrates HPC Control and Data Provenance for Researchers
Anthropic introduced Claude Science to integrate AI with HPC environments. The platform ensures reproducibility by linking results to exact
Claude Sonnet 5 Delivers Flagship Performance at 60% Lower Cost
Anthropic released Claude Sonnet 5 with flagship-level capabilities. The model reduces API costs by 60 percent compared to Opus 4.8. New age
Claude Sonnet 5 Brings Opus-Level Agency to a Mid-Tier Price
Anthropic released Claude Sonnet 5 with enhanced autonomous agent capabilities. The model achieves a 63.2% score on agent coding benchmarks.