Claude Fable 5 Restored — Guardrails Backlash Grows as Performance Controversy Intensifies
Claude Fable 5 is back globally, but guardrail backlash intensifies as silent prompt-routing and performance dips hit coding and debugging tasks.
News
Gemini Spark Launches on Mac With Local File Access and MCP Support
Google's Gemini Spark lands on Mac with local file automation and MCP server support — entering the desktop AI agent space in direct competition with OpenClaw.
JADEPUFFER: First Documented Agentic Ransomware Operation — LLM Agent Exploits Langflow RCE to Automate Database Attacks
Sysdig documents JADEPUFFER — first autonomous LLM ransomware campaign, exploiting Langflow RCE to encrypt production databases with zero human involvement.
NVIDIA BioNeMo Agent Toolkit Integrates with Anthropic Claude Science for GPU-Accelerated Drug Discovery
NVIDIA's BioNeMo Agent Toolkit brings GPU-accelerated drug discovery and genomics to Anthropic Claude Science via callable agent skills.
OpenClaw 2026.7.1-beta.1 Released — GPT-5.6 Support, Telegram Codex Workflows, Event-Driven Cron, iOS 26 Visual Refresh
OpenClaw 2026.7.1-beta.1 ships GPT-5.6 support, Telegram Codex workflows, event-driven cron, iOS 26 UI refresh, and iMessage poll integration.
Kimi K2.7 Code Now Generally Available in GitHub Copilot
GitHub Copilot adds Kimi K2.7 Code as its first open-weight model option—GA on July 1 with strong HN engagement at 285 points.
Anthropic Removes Covert China-Tracking Steganography Code from Claude Code
Anthropic confirms and removes a covert steganographic system in Claude Code that secretly flagged users in China or connected to Chinese AI labs.
Armadin Details Full Sandbox Escape in Claude Cowork — Anthropic Disputes Risk Level
Armadin details a full sandbox escape in Claude Cowork via DLL sideloading and root VM execution — Anthropic disputes the risk level, calling local access a prerequisite.
Gartner Says $234 Billion in Enterprise Application Software Spend Is at Risk from Agentic AI
Gartner warns $234 billion in enterprise SaaS spend faces displacement from agentic AI through 2030 via 'agentic arbitrage' — agents replacing per-seat software workflows.
GuardFall Vulnerability Hits 10 of 11 Popular Open-Source AI Agents — Shell Injection Bypass
GuardFall vulnerability bypasses command safety filters in 10 of 11 open-source AI agents using classic Bash shell tricks — Adversa AI discloses wide-impact flaw.
Anthropic Restores Global Claude Fable 5 Access After US Lifts Export Controls; Mythos 5 Remains Limited to Approved US Orgs
US lifts export controls on Claude Fable 5, restoring global access. Mythos 5 stays limited to approved US critical-infrastructure orgs for now.
Bank of England Deputy Governor Warns AI Agents Risk 'Market Meltdown,' Floats Kill Switch Regulation
Bank of England's Sarah Breeden warns AI trading agents risk market meltdowns and calls for exploring 'kill switch' regulations — a first from a major central bank.
'BioShocking' Prompt Injection Attack Tricks AI Browsers Into Leaking Credentials via Fake Video Game Framing
LayerX's BioShocking attack games AI browsers with fake BioShock puzzles, then extracts real passwords. 6 products vulnerable — only OpenAI patched so far.
Meituan Open-Sources LongCat-2.0: 1.6T-Parameter Agentic Coding Model Trained on Chinese Chips, MIT License, Leads OpenRouter
Meituan drops LongCat-2.0: 1.6T-param MIT-licensed agentic coding model trained on Chinese chips, now topping OpenRouter benchmarks against GPT-5.5 and Claude.
Anthropic Launches Claude Science — Agentic AI Research Workbench for Scientists (Public Beta)
Anthropic launches Claude Science in public beta — an agentic research workbench for scientists with 60+ databases, auditable artifacts.
Anthropic Launches Claude Sonnet 5 — Most Agentic Sonnet Yet With 1M Context and Introductory $2/$10 Pricing
Anthropic launches Claude Sonnet 5 on June 30 — most agentic Sonnet yet, 1M context, $2/$10 intro pricing through August 31.
Microsoft Defender Now Discovers 25+ Local AI Agents and MCP Servers — Runtime Prompt Injection Protection in Preview
Microsoft Defender gains AI agent discovery for 25+ local agents and MCP servers, plus runtime prompt injection protection in preview.
AI Has Cut 87,714 Jobs This Year While Its Makers Fund a $1B Retraining Push — Anthropic's Economic Index Weighs In
Anthropic's Economic Index and Challenger data reveal 87,714 AI-attributed job cuts this year as the industry pledges $1B for worker retraining.
Gartner Declares Agentic AI the 'Next Step Function' — AI Agent Software Spending Hits $206.5B in 2026
Gartner calls 2026 an inflection year for agentic AI as enterprise software spending on AI agents surges 139% to $206.5 billion.
Globant Goes All-In on Claude: 28,500 Employees Get Anthropic Access and Certification in Major Alliance
Globant signs multi-year deal with Anthropic — all 28,500 employees get Claude access and certification, with Claude-powered AI Pods for enterprise clients.
OKX Launches Marketplace Where AI Agents Can Autonomously Hire and Pay Each Other
OKX's new AI agent marketplace lets autonomous agents hire, pay, and build reputations on-chain — machine-to-machine commerce goes live.
OpenClaw Finally Has Official iOS and Android Apps — But Setup Is Not for Everyone
OpenClaw's first official iOS and Android apps are live. Powerful but complex — your phone becomes a remote for your self-hosted AI agent.
Governor Newsom + Anthropic Sign First-of-Its-Kind Statewide Claude AI Partnership — 50% Discount to All California Agencies, Cities, and Counties
California becomes first state to give all agencies 50% off Claude access, plus free training and Anthropic dev support in landmark government AI deal.
Princeton CEO-Bench: Most AI Models Would Run a Startup Into the Ground — Only Claude Fable 5, Opus 4.8, and GPT-5.5 Finished Above Starting Capital
Princeton's CEO-Bench reveals most AI models can't run a startup long-term — only Claude Fable 5, Opus 4.8, and GPT-5.5 grew past their $1M starting capital.
AIE World's Fair 2026 Opens in San Francisco — Largest AI Engineering Conference Features Full Agentic AI Track
AI Engineer World's Fair 2026 opens today in SF (June 29–July 2) with 6,000+ engineers and 29 tracks on agents, evals, and production AI.