🔥 Fable 5 is back: A major shift for every AI team.

Advertise here | 6-min Read

Want to get the most out of ChatGPT?

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide

Fable 5 is back — a major shift for every AI team.

The US Department of Commerce lifted its export control directive on June 30 at 3:31 PM ET. Fable 5 returned globally on July 1 across Claude.ai, Claude Platform, Claude Code, and Claude Cowork. The resolution was not a policy change. It was a proof problem — and the proof went against the government.

The details:

• The jailbreak that triggered the June 12 ban — a prompt that caused Fable 5 to identify software vulnerabilities — was subsequently tested against Claude Opus 4.8, GPT-5.5, and Kimi K2.7. All three reproduced the same exploit. Fable 5 had no unique offensive capability the government needed to contain.

• Anthropic is now co-developing, with Amazon, Microsoft, and Google, a framework for scoring how dangerous a given jailbreak actually is — to prevent future situations where a borderline finding triggers a disproportionate government response.

• In its July 1 statement, Anthropic made a direct policy ask: "Government involvement in AI releases requires a durable, transparent process that gives cyber defenders and others the certainty they need about access to powerful models. These rules should be codified in strong regulation and applied equally across frontier model developers.

  Why it matters: The 19-day Fable 5 ban will be studied for years. Not because of what it did to Anthropic's revenue — though that was significant but because of what it revealed about the governance gap at the frontier.

AI CHEAT SHEET

AI news highlights

• Only 57,000 jobs added in June — lowest since 2024 — and AI is the named cause — tech layoffs total 142,000 YTD. RAISE US: 88,000 US job cuts directly attributed to AI in 2026.
• White House voluntary AI standards announcement expected next week — August 1 hard deadline — replaces bilateral deals with a framework. Both OpenAI and Anthropic have asked for exactly this. .
• Anthropic's AI for science finding: deterministic tools push AI biology accuracy from 16.9% to 92.8% — the gap between Claude alone and Claude plus structured tool use. The most consequential research result Anthropic published in June.
• Grok 4.3 is now on Amazon Bedrock at $2.50 output per million tokens — substantially cheaper than Claude Sonnet 5 at $10 or GPT-5.5 at $15. Frontier-adjacent at a fraction of frontier pricing.
• Google Search is now entirely powered by Gemini 3.5 Flash — custom AI-summarised pages replace traditional link lists for all queries. The largest single AI deployment in consumer search history.
• Claude Sonnet 5 standard pricing kicks in September 1: $3 input / $15 output — introductory $2/$10 ends August 31. New tokenizer means effective cost may be 10-35% higher for some workloads despite the same nominal rate.

Write a newsletter that people actually look forward to

Prompt: You are a newsletter editor who has grown publications from zero to hundreds of thousands of subscribers — not by posting more, but by making each issue feel like something the reader would be annoyed to have missed. You know that the inbox is the most competitive attention environment on the internet, and you treat every issue like a product launch.

Here is what I want to write about this week: [describe your topic, the key stories or insights you want to cover, and any specific piece you want to anchor the issue around].

My audience: [describe your readers — their role, what they care about, how sophisticated they are on this topic].

Build me a full issue across five parts:

1. The subject line — Write three options. The first should create curiosity without being clickbait. The second should use a specific number or data point. The third should feel like something a colleague would forward with "you need to read this." For each, tell me why it works and who it targets.

2. The opening — Write the first three sentences of the issue. This is all most people will read before deciding whether to continue. Do not open with "this week in..." or a summary. Open with the tension, the implication, or the thing that makes this week different from every other week. Make them need to keep reading.

3. The main story — Write the centrepiece section: the most important development this week, why it matters to this specific reader, and what they should do about it. Not a news report. An interpretation. What does this mean for someone in their role, with their constraints, making decisions on their timeline?

4. The quick hits — Five bullet points: each one a different story, under 30 words, that tells the reader something they did not know and makes them want to know more. The goal is not to summarise — it is to create a reason to click.

5. The close — One paragraph that ends the issue in a way the reader will remember. Not a summary. Not "see you next week." A thought, a question, or an observation that makes them think about this topic differently after they close the email.

One rule: every sentence must earn its place. If a sentence does not add information, create tension, or move the reader forward — cut it.

America’s Founding, Rewritten in Google Workspace

Trending AI tools

• Retrace — Debug AI agents by replaying and forking runs — launched 7 hours ago
• Scritty — Shared, searchable memory for every AI coding agent — launched 7 hours ago
• Acti — Agentic keyboard for mobile commands and search
• Backgrind — Run parallel AI coding agents over any app, sandboxed with worktrees
• Deep Work Plan — Repo-native specs and resumable tasks for longer agent refactors
• Wispr Flow — Dictate in any app — writes in your voice, not generic AI prose
• Framer Agents — Design, write, and organise your site with agents — Claude Code and Codex compatible
• Granola — AI meeting notes on top of any call tool — no bot, no permissions needed

That’s a Wrap

SPONSOR US

Get your business in front of over 90k+ AI professionals

8020AI is the world’s #1 AI Newsletter, Read by 90k+ professionals from leading companies such as Google, OpenAI, Meta, and Microsoft.

We've assisted in promoting Over 500 AI-Related Products. Will yours be the next?

What We Can Offer:

• Launch an Advertising Campaign

• Introduce New Product or Features

• Other Business Cooperation

Get in touch today.

Or Email our founder Alamin at [email protected]

FEEDBACK

How was your experience with 8020AI today?

Login or Subscribe to participate in polls.

If you have specific feedback or anything interesting you’d like to share, please let us know by replying to this email.