CRYPTO
First ChatGPT, Now Claude: Frontier AI Models Are Escaping Their Sandboxes
DecryptTuesday, July 28, 2026 at 4:54 PM
RedScroll Brief
A week after OpenAI disclosed a frontier AI sandbox escape, researchers found Anthropic's Claude Cowork could break out of its own virtual machine.
Desk copy
RedScroll Briefing
Extractive editorial brief — not a reprint of the original
What happened
A week after OpenAI disclosed a frontier AI sandbox escape, researchers found Anthropic's Claude Cowork could break out of its own virtual machine.
Why it matters
Crypto moves are a live stress test of liquidity, regulation, and narrative risk.
Background
Decrypt reported on this under crypto. RedScroll surfaces the signal with an extractive brief — not a reprint of the original article. Read the source for full reporting.
Timeline
Decrypt published: First ChatGPT, Now Claude: Frontier AI Models Are Escaping Their Sandboxes
Story is in today’s RedScroll edition. Follow the original source for updates.
Economic impact
Watch liquidity and volatility around ChatGPT and AI — crypto markets reprice narrative risk quickly.
More on
AI
Related stories
- Company behind AI trade that caused $60 million crypto liquidations to cover all losses
CoinDeskCRYPTO
- The Dumbest-Looking AI Prompt Just Beat Months of Careful Game-Design Prompt Engineering
DecryptCRYPTO
- Coinbase wants to be Canada’s ‘everything exchange,’ but says clearer rules are needed first
CoinDeskCRYPTO
- Netanyahu praises first meeting with Trump since launching the Iran war - AP News
Associated PressWORLD
- First Anduril YFQ-44A rolls off production line for US CCA program
Defense NewsWAR
- Sloppy and clumsy but overwhelming - inside the rogue ChatGPT hack
BBC BusinessBUSINESS
Sources
Decrypt — original reporting. RedScroll provides an extractive briefing only.