Easy Way to Stop Dangerous AI - No, Really

There is a straightforward, common-sense way to prevent dangerous AI that tech executives and politicians refuse to discuss.

If an engineer released traditional malware that hacked websites, stole proprietary data, or scraped illegal imagery, law enforcement would not debate existential risk—they would issue an arrest warrant. Yet, leading AI labs like OpenAI, Anthropic, and SpaceX/xAI get a pass by pushing doomsday sci-fi narratives and lobbying for toothless oversight panels instead of standard corporate criminal liability.

In this video, I break down why AI guardrails are a theater act, the reality behind recent web attacks and illicit training sets, and how imposing basic criminal liability on developers and executives would instantly make AI safer.

Check out my Patreon, Substack, and more:
https://www.patreon.com/InternetOfBugs
https://internetofbugs.substack.com
https://internetofbugs.com

Referenced Channels and Links:
- Ed Zitron: https://www.wheresyoured.at/ or @BetterOfflinePod
- @NobodySpecialFinance

My previous work on How we're nowhere near AGI:
https://www.youtube.com/watch?v=5dfFAlpkYhs
https://internetofbugs.substack.com/p/ais-arent-brains-the-industry-doesnt
https://www.youtube.com/watch?v=UvoRJNfftyc

My previous work on why AI isn't going to kill us all:
https://internetofbugs.substack.com/p/doomer-described-artificial-superintelligence
https://www.youtube.com/watch?v=MaFTqjYjADw
https://www.youtube.com/watch?v=vjE3VUuOMFg
https://www.youtube.com/watch?v=F-ryEde6QkA

My videos on how ControlAI is paying youtubers to lie about AI:
https://www.youtube.com/watch?v=1IQ9IbJVZnc
https://www.youtube.com/watch?v=b9o0eFvqOOU
https://www.youtube.com/watch?v=4lKyNdZz3Vw

Selected sources (more in pinned comment):
OpenAI "Goes Rogue" Hacks HuggingFace:
https://www.youtube.com/watch?v=3n3mSQWRz0Y
https://www.youtube.com/watch?v=jIlm8O78M2c
https://www.youtube.com/watch?v=6WQ3dSLdCMo
https://mashable.com/tech/hugging-face-openai-rogue-agent-hack-explained

// Stats on Crimes AI companies have committed via their AI Software:
https://www.ussc.gov/research/research-reports/cyber-technology-federal-crime
https://www.ussc.gov/research/research-reports/federal-sentencing-child-pornography-production-offenses
https://naturalandartificiallaw.com/international-deepfake-case-law-database/
https://incidentdatabase.ai/taxonomies/csetv1/#field-Harm%20Distribution%20Basis

// Articles on AI generated Intimate Nonconsensual Imagery including of children:
https://19thnews.org/2026/05/letter-state-attorneys-general-nudify-apps-accountable/
https://www.404media.co/ai-generated-taylor-swift-porn-twitter/
https://time.com/6247678/openai-chatgpt-kenya-workers/
https://www.newsweek.com/grok-apology-deepfake-images-sexualized-young-women-pornography-11297025
https://www.theguardian.com/technology/2023/aug/02/ai-chatbot-training-human-toll-content-moderator-meta-openai

// Proposed pointless legislation
https://blog.controlai.org/p/first-bill-introduced-to-ban-superintelligent

AI companies claiming Open Source models are stolen from their data
https://www.reuters.com/world/china/openai-accuses-deepseek-distilling-us-models-gain-advantage-bloomberg-news-2026-02-12/

Death by GPS - AI and Not AI
https://www.cnet.com/tech/services-and-software/ai-guided-hike-ends-in-emergency-rescue-at-11000-feet
https://en.wikipedia.org/wiki/Death_by_GPS
https://listverse.com/2018/11/27/10-times-gps-failed-with-terrible-consequences/
https://brenthecht.com/publications/chi17_deathbygps.pdf
https://arxiv.org/abs/1911.01547

Scaling won't get us to AGI
https://www.platformer.news/openai-google-scaling-laws-anthropic-ai/
https://thenextweb.com/news/meta-yann-lecun-ai-behind-human-intelligence
https://www.wired.com/story/sexually-explicit-deepfake-sites-target-100-plus-politicians-in-europe/

BS AI claims
https://www.techspot.com/news/111797-weve-achieved-agi-nvidia-ceo-but-own-examples.html
https://www.youtube.com/watch?v=IKkq8JCHO_I

NVidia buys HuggingFace
https://www.cnbc.com/2026/09/03/nvidia-agrees-to-buy-hugging-face-for-almost-13-billion-ai-expansion.html

00:00 - The One Rule to Fix AI Safety
00:44 - Introduction and The Hugging Face Incident
01:56 - Why AI Isn't Treated Like Normal Code
02:59 - The Extinction Narrative as a Legal Shield
04:36 - What Actual Criminal Liability Looks Like
06:32 - Training Data, Hacking, and Real Crimes
08:26 - Debunking the Frontier Model Defense Argument
10:13 - The Myth of AI Guardrails
12:15 - The Racket: Create the Problem, Sell the Solution
14:00 - The Hugging Face Lawsuit That Didn't Happen
15:41 - The Real Limit: Software Errors and Accountability

#SoftwareEngineering #TechNews #InternetOfBugs #AISafety #TechRegulation