31 October 2026
Claude Code extension on Cursor was backdoored, while AI spending fears hit markets
A backdoored Claude Code extension on the Cursor marketplace was the main security story today. AI companies also agreed a shared way for websites to opt out of AI training. Investors grew nervous about the cost of AI, and HighLevel added usage limits for SaaS clients.
Security and safety
Anthropic VS Code extension on Cursor backdoored for nearly 12 hours (opens the source in a new tab)
Ars Technica reports the Anthropic extension on Cursor carried a backdoor for nearly 12 hours. It is still unknown how attackers got into Anthropic's verified Cursor account.
Why it matters: If you or your developers installed Claude Code in Cursor, check which version you have and watch for official advice from Anthropic.
Ars Technica · 30 Oct
Backdoor in Claude Code for VS Code extension on the Cursor Marketplace (opens the source in a new tab)
A security write-up describes a backdoor in the Claude Code extension for VS Code, released on the Cursor Marketplace. It drew heavy discussion on Hacker News.
Why it matters: Treat editor extensions like any other software your team installs, and ask your developers to review what they have added.
phorion.io via Hacker News · 30 Oct
OpenAI will honor shared signal for AI training preferences (opens the source in a new tab)
OpenAI says it will honor a shared robots.txt signal that lets website operators state their AI training preferences to participating providers.
Why it matters: If you run a website or publish content, you may soon be able to set your AI training preference once instead of company by company.
OpenAI · 30 Oct
Big AI firms back new way to opt out of scraping (opens the source in a new tab)
OpenAI, Google, Microsoft and Anthropic say they will adopt a new standard for publishers to opt out of AI scraping. A shared "policy token" signal lets publishers set preferences across several AI companies at once.
Why it matters: Ask whoever manages your website to look at this control if you do not want your content used for AI training.
The Verge · 30 Oct
Tools and products
HighLevel adds usage limits for SaaS clients (opens the source in a new tab)
You can now set a usage limit for each SaaS client and choose what happens at the limit, such as billing for extra use or stopping the action. Limits start with Voice AI and Conversation AI.
Why it matters: If you resell HighLevel AI features, you can stop surprise costs and charge clients when they use more.
HighLevel · 30 Oct
HighLevel introduces new web templates (opens the source in a new tab)
HighLevel announced new templates for websites, sales funnels and email campaigns. It says they are fully customizable.
Why it matters: If you build sites, funnels or emails in HighLevel, look through the new templates to save setup time.
HighLevel · 31 Oct
OpenAI adds ChatGPT safeguards and teen protections (opens the source in a new tab)
OpenAI says it is expanding violence prevention safeguards in ChatGPT, with earlier interventions and a new commitment to public safety reporting.
Why it matters: If your business uses ChatGPT with clients or young users, expect more safety checks and reporting from OpenAI.
Automation
n8n explains throttling, rate limits and retry backoff (opens the source in a new tab)
An n8n guide covers how Retry-After, exponential backoff and jitter make retries more reliable when an API limits requests.
Why it matters: If your n8n workflows fail when an app says "too many requests", this guide shows how to make them retry safely.
n8n · 30 Oct
Cisco gives engineer unlimited AI tokens to rewrite legacy code (opens the source in a new tab)
Fortune reports that Cisco gave an engineer unlimited AI tokens to rewrite legacy code.
Why it matters: Old systems are a common target for AI help, so a small team could test AI on one old tool of its own.
fortune.com via Hacker News · 30 Oct
Few people use AI agents for DoorDash orders (opens the source in a new tab)
Ars Technica reports that basically no one is using AI agents for DoorDash orders. It says this e-commerce use case is still early.
Why it matters: If you plan to let customers buy through AI agents, expect low use for now.
Enterprise
Cracks are emerging in the AI spending boom (opens the source in a new tab)
TechCrunch says Big Tech is spending billions on AI infrastructure, but debt, negative cash flow and investor doubt are raising questions about how long it can last.
Why it matters: Prices and plans for AI tools could change if the companies behind them face funding pressure, so avoid locking into one vendor.
TechCrunch · 30 Oct
Wall Street closes lower on AI cash burn fears (opens the source in a new tab)
Reuters reports a sharp fall on Wall Street as fears about AI cash burn sparked a tech selloff.
Why it matters: Watch for price rises or plan changes from AI vendors, and keep your tools easy to switch.
reuters.com via Hacker News · 30 Oct
OpenAI raises less than half of $30B target (opens the source in a new tab)
The Wall Street Journal reports that OpenAI raised less than half of its $30B target.
Why it matters: If you build on OpenAI, keep a backup model option in case its finances change what it offers.
Also worth knowing
Anthropic cracking down on video game rewrites (opens the source in a new tab)
A post on X, shared on Hacker News, says Anthropic is cracking down on video game rewrites.
Why it matters: Check Anthropic's usage rules before using Claude to recreate existing software or games.
twitter.com via Hacker News · 30 Oct
Apple Intelligence coming to China with iOS 27.3 (opens the source in a new tab)
Apple says Apple Intelligence will reach China with iOS 27.3. A beta is due on November 4th and the full release is set for December.
Why it matters: If you have clients or staff in China, Apple's AI features may soon be available to them.
The Verge · 30 Oct
Privacy-focused AI assistant Ollie is shutting down (opens the source in a new tab)
TechCrunch reports that the founder of Ollie, an AI assistant that respected user privacy, says it will shut down in a week.
Why it matters: If you use small AI tools, keep copies of your data and have a fallback in case they close.
TechCrunch · 30 Oct