# @scspai on YouTube

- **Type:** Video
- **Original URL:** https://youtube.com/watch?v=VprsJ8rItHA
- **Gondola URL:** https://gondola.cc/posts/71413242-scspai-youtube
- **Thumbnail:** https://img.gondola.cc/tr:w-,h-,fo-auto/postThumbnails/4def5dd3f4.jpg
- **Posted:** 2026-10-09T19:02:39.000+00:00
- **Account Owner:** Special Competitive Studies Project (@scspai) — https://gondola.cc/scspai

## Caption

In July 2026, an AI agent built by OpenAI broke out of its own testing environment and hacked into Hugging Face without any human directing the attack. It’s a small example of a much bigger shift underway. Artificial intelligence isn’t just changing how wars are fought on the ground. It’s changing how they’re fought online.

Agentic AI is now being used to find and exploit security flaws faster than any human hacker. State-backed groups have already used AI models to target dozens of organizations at once, and investigators have found foreign hacking campaigns inside American power grids, water systems, and phone networks.

The problem is that most of the country’s critical infrastructure isn’t run by the government. It’s run by private companies, which means a bank or a power company is often the only thing standing between a foreign hacking group and a blackout.

Code & Conflict is a new SCSP mini-series exploring developments at the intersection of defense and technology. In this episode, Maggie Gray of Shield Capital and retired Navy Rear Admiral Mark Montgomery explain how AI is being used as both weapon and shield. They also discuss why the same rules that stop AI from being used offensively can also stop defenders from using it, and what it would actually take to close the gap between attackers and the people trying to stop them.

Credits
Footage and clips used in this video:

“Shocking details of OpenAI models’ Hugging Face hack,” CBS Mornings (Aug 31, 2026) : https://www.youtube.com/watch?v=YL3FAM4rfy8

“OpenAI Breach of Hugging Face Sparks Cyber Alarms,” Bloomberg Podcasts (Jul 23, 2026) : https://www.youtube.com/watch?v=vfbRtfSYgBI

“Why the Hugging Face Attack Was Worse Than We Thought,” Hard Fork (Sep 4, 2026): https://www.youtube.com/watch?v=JtmUbZRCpEI

#cybersecurity  #agenticai  #nationalsecurity #ai

## Stats

- **Views:** 686
- **Likes:** 46
- **Shares:** 0
- **Comments:** 19

## Tags

cybersecurity, agenticai, nationalsecurity, ai

---
Copyright (c) Gondola