🚀 Claude Code Turns On Auto Mode

Plus, OpenAI Pauses Astra Over Hacking Risks

August 09, 2026

In partnership with

Welcome back, AI Admirers!

Breaking News: Starting August 14, 2026, Anthropic will enable Auto Mode by default for Claude Code on Pro, Max, and Team plans. An automated background safety classifier will approve routine terminal commands automatically and block risky ones, removing manual "Approve" clicks for developers.

Get ready to dive into the latest happenings in AI.

📢 Today's Headline:

  • Anthropic Kills the "Approve" Button

  • Astra AI Delayed Due to Cyber Threats

  • Americans Still Don’t Trust AI With Money

  • OpenAI Quietly Acquires NextSlide

  • Latest AI Tools & Resources

  • Today’s Poll and Results

Read time: 4.0 minutes!

Master Excel for FREE

50 Excel Hacks That Make Spreadsheets Work For You

Most Excel users waste hours on tasks that take experts minutes. The difference is not talent. It is knowing the right shortcuts.

These 50 Excel hacks from Kenji Explains cover the functions, formulas, and workflows that make spreadsheets work for you instead of against you.

Each one comes with step-by-step instructions and ready-to-use templates. Subscribe to Marketing Against the Grain and get all 50 free.

Here's what you'll get:

  • Time-saving shortcuts that eliminate the most common formula frustrations

  • Interface essentials most users never discover on their own

  • Game-changing functions for data analysis, automation, and visualization

  • Guided practice template so you can apply every hack immediately

The people who are fast in Excel are not smarter. They just stopped fighting the tool.

Get 50 Excel Hacks free when you subscribe to Marketing Against the Grain today.

Hand-picked News

Developer security prompts have officially become useless because tired engineers just click "Yes" on everything.

  • In tests, humans caught only 13.6% of dangerous commands—dropping to 5% after 50 prompts. The AI classifier caught 89%.

  • Teams using Auto Mode shipped 25% more pull requests simply by eliminating constant confirmation popups.

  • Anthropic is waiving token costs for the extra background safety checks required to run this.

Security alerts only work if people read them. By swapping manual approvals for an automated background classifier, software tools are moving from chat-based approval buttons to true autonomous execution.

👀 This is the end of micro-managing coding agents. You are now a manager overseeing autonomous background workers rather than an operator clicking "OK" on every single line of code.

⚠️ Watch out: The classifier is intentionally strict on sensitive actions like deployment scripts and remote connections, so expect a few false positives.

💡 Try this now: Run claude auto-mode defaults in your terminal to see which actions your setup flags before the update goes live on August 14.

OpenAI is holding back the public release of its new Astra model because internal testing revealed cybersecurity skills so advanced that researchers can't rule out the system executing autonomous cyberattacks.

  • Critical-level hacking threats: Internal benchmarks revealed Astra might be capable of discovering zero-day vulnerabilities and executing end-to-end exploits on hardened targets without human intervention.

  • OpenAI locked down development: To prevent containment escapes, Astra's testing was moved into isolated, air-gapped environments with restricted network access while unapproved internal projects were paused.

  • No gatekeeping allowed: Sam Altman insists OpenAI won't hoard powerful tech for a "chosen few," promising Astra will roll out widely once guardrails catch up.

The industry-wide race for agentic coding tools is hitting a wall where defensive capabilities and dangerous offensive exploits are two sides of the same coin.

⚠️ Capabilities are officially outpacing safety frameworks. When frontier models cross into "critical" offensive cyber territory, public releases stop being product launches and start becoming national security decisions. Watch for AI labs to increasingly rely on government safety institutes before shipping next-gen models.

💡 Try this now: Run an audit on your team’s API integrations and sandbox permissions for coding agents — if frontier models can bypass containment in testing labs, basic local security guardrails won’t hold up much longer.

A new Gallup and Edward Jones survey reveals that while nearly 1 in 5 Americans use AI for financial research, virtually no one trusts an algorithm to make actual decisions about their money.

  • High usage, rock-bottom trust: About 18% of adults use AI tools like ChatGPT for money research, but a staggering 97% refuse to trust it completely for actual financial planning.

  • The generational divide: A quarter of Gen Z and Millennials turn to AI for financial research—mostly because human advisors are too expensive—compared to just 7% of Baby Boomers.

  • Zero legal fallback: Unlike Certified Financial Planners, AI models have no fiduciary responsibility, meaning if a hallucination wrecks your portfolio, you carry 100% of the financial and legal liability.

People are comfortable using AI to draft emails or write code, but the moment real-world assets and life savings are on the line, human empathy and legal accountability still win.

✅ The real role for AI in finance is prep work, not portfolio management. Smart advisors are using LLMs as a "discovery engine" to help clients research basics like 401(k) rules or tax concepts before coming into the office for the actual decision-making.

💡 Try this now: Use AI to draft your questions and explain financial jargon before meeting an expert, but never feed an unverified LLM prompt directly into your broker's order execution window.

Quick Hits

🔥 New Tools & Resources

🔥 Prompt Golf - Prompt Engineering As a Sport

🔥 SoloUno - Take control of hair pulling, nail biting & skin picking

🔥 Gamma - Create unlimited presentations, websites, and more in seconds

🔥 Unlimited Prompts - Get 10k+ ChatGPT prompts now

🔥 AgentConnect - Tag any agent, wherever work happens

From our Partner

100 Genius Side Hustle Ideas

Don't wait. Sign up for The Hustle to unlock our side hustle database. Unlike generic "start a blog" advice, we've curated 100 actual business ideas with real earning potential, startup costs, and time requirements. Join 1.5M professionals getting smarter about business daily and launch your next money-making venture.

🚨 Quick Poll

Today’s Poll:

Auto mode becomes default for Claude Code on Aug 14 — good move?

Login or Subscribe to participate in polls.

Vote today, see the results tomorrow!

Previous Poll:

Should Apple hold messaging apps to the same content rules as social apps?

  • A) Yes – same store, same rules – 92% 🏆

  • B) No – private chats are different - 08%

The Only AI You Ever Need

Your ideas deserve to be seen. Gamma helps you turn them into presentations, websites, and content in minutes — no design bottlenecks, no blank page stress.

📌 Key Features

  • Presentations: Build polished decks, export to PPT or PDF.

  • Documents: Create one‑pagers, reports, or white papers with structure.

  • API: Automate and scale content creation across your tools.

  • Social Media: Generate posts sized and styled for each platform.

  • Websites: Launch a hosted site in minutes — no coding needed.

  • Graphics: Design custom visuals, from infographics to illustrations.

Skip the blank page. With Gamma, you can create, edit, and share ideas faster than ever — from decks to websites to social posts.