Скачать или смотреть Your AI Coding Assistant is a SECURITY RISK

In this video I dive into the hidden security threat lurking in AI coding assistants and the 'vibe coding' lifestyle. Malicious actors are exploiting AI coding by creating malware packages with names commonly hallucinated by LLMs.

You don't need another AI tutorial. You need a room full of people who are actually building!
Join the AI Captains Academy: skool.com/aicaptains (free)

Ready to go deeper? Free 5-day email courses:
→ AI Foundations Sprint: https://foundations.aicaptains.ai
→ Command Line Bootcamp: https://bootcamp.aicaptains.ai

I use this as a chance to compare basic app prototyping with Firebase Studio, Bolt, and Cursor (via Claude).

This is the article I refer to in the video: https://www.bleepingcomputer.com/news...

It's happening more in open-source models, but according to the article, even GPT is still prone to hallucinating package risks.

🔧 Tools used:
Cursor AI (https://cursor.sh)
Claude Sonnet 3.7 (https://claude.ai)
Bolt (https://bolt.new)
Firebase Studio (https://firebase.studio)
Venice AI: https://venice.ai/chat?ref=PA5RHk

⏱️ Timestamps:
0:00 - Introducing the AI Security Threat
2:22 - Most Vulnerable AI Models vs Safer Options
3:02 - Building Verification Tools: Firebase Studio vs Alternatives
6:13 - Advanced Security: Signature and Hash Verification
7:24 - Live Testing All Solutions with Real and Fake Packages
14:36 - The Winner

Follow my build-in-public journey in my personal newsletter: https://blog.jordanurbs.com

Say hi on X: https://x.com/jordanurbs