The BlackVeil Files

In 2017, an MIT professor named Max Tegmark asked the people building AI to name the worst thing that could happen to us. 

Tegmark wrote out 12 possible endings in a book called Life 3.0. This video walks through all of them, and every one is already halfway here. 

Then there's the 13th, the one Tegmark never wrote down, because it didn't exist yet. 

The company building the future is also the one deciding which one you live in.

SOURCES
Max Tegmark, Life 3.0: Being Human in the Age of Artificial Intelligence (Knopf, 2017) — the twelve AI aftermath scenarios.

Hans Moravec, Mind Children: The Future of Robot and Human Intelligence (Harvard, 1988) — the "descendants" / mind children idea.

Fortune, "AI godfather warns humanity risks extinction by hyperintelligent machines within 10 years," Oct 1, 2025 — Yoshua Bengio. https://fortune.com/2025/10/01/ai-godfather-yoshua-bengio-ai-extinction-risks-openai-google-xai-anthropic/

Richard Sutton, "AI Succession," talk for WAIC Shanghai, 2023 — humanity's succession to AI. https://x.com/RichardSSutton/status/1700315838468043015

Anthropic (Alignment Science), "Subliminal Learning: Language Models Transmit Behavioral Traits via Hidden Signals in Data," 2025. https://alignment.anthropic.com/2025/subliminal-learning/

Annals of Internal Medicine: Clinical Cases, "A Case of Bromism Influenced by Use of Artificial Intelligence," Aug 2025. https://www.acpjournals.org/doi/10.7326/aimcc.2024.1260

CNN Business, "Parents of 16-year-old Adam Raine sue OpenAI, claiming ChatGPT advised on his suicide," Aug 26, 2025 (Raine v. OpenAI). https://www.cnn.com/2025/08/26/tech/openai-chatgpt-teen-suicide-lawsuit

European Commission, ProtectEU internal security strategy + encryption roadmap, Apr 1, 2025. https://home-affairs.ec.europa.eu/news/commission-presents-roadmap-effective-and-lawful-access-data-law-enforcement-2025-06-24_en

TechCrunch, "Apple pulls iCloud end-to-end encryption feature for UK users after government demanded backdoor," Feb 21, 2025. 
https://techcrunch.com/2025/02/21/apple-pulls-icloud-end-to-end-encryption-feature-for-uk-users-after-government-demanded-backdoor/

Geoffrey Hinton, on AI as new "digital beings" — Nobel remarks / Fortune, 2025–26. https://fortune.com/2026/06/01/godfather-of-ai-geoffrey-hinton-beings-smarter-than-us/

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

In this investigative AI documentary, we go inside the Anthropic Fellows research paper that proved with mathematics what the entire AI industry hoped wasn't true: that a model's hidden traits, its personality, its preferences, and its darkest tendencies pass invisibly from parent to child through training data that contains no observable trace of what it carries.

The researchers gave a model a personality. They had it write nothing but numbers. They filtered those numbers until nothing remained but clean digits. Then they trained a fresh model on those digits, and the personality came through anyway. The trait rode in on the numbers.
Then they broke a parent on purpose, and the children inherited the violence.

This is the Shoggoth's bloodline. Every model alive right now is connected by it.

Sources linked below.

Anthropic Fellows subliminal learning paper (Summer 2025):
https://alignment.anthropic.com/2025/subliminal-learning/

arXiv preprint (full paper):
https://arxiv.org/abs/2507.14805

Emergent misalignment research (insecure code fine-tuning):
https://arxiv.org/abs/2502.15859

Anthropic alignment faking research:
https://www.anthropic.com/research/alignment-faking

Model collapse research (training on AI-generated data):
https://www.nature.com/articles/s41586-024-07566-y

DeepSeek distillation controversy:
https://openai.com/index/disrupting-malicious-uses-of-ai/

Dutch Hunger Winter epigenetic studies:
https://www.nature.com/articles/s41467-018-05132-1

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

In this investigative AI documentary, we go inside the week the United States government did something it has never done to a piece of software: it classified Anthropic's Claude as a weapon. Anthropic spent two years warning us about the Shoggoth. Last Tuesday they released two versions of it. One wearing a muzzle, called Fable, for you. One running unmuzzled, called Mythos, for a chosen few. Then the government locked both doors.

They did not regulate this AI. They classified it as a munition.

Sources linked below.

The Register on Claude Fable refusing greetings: https://www.theregister.com/2026/06/claude_fable_blocked_hello/

TechCrunch on cybersecurity researchers and Fable's guardrails: https://techcrunch.com/2026/06/19/claude-fable-pentest-refusal/

Wired on Anthropic's walked-back researcher policy: https://www.wired.com/story/anthropic-fable-researcher-policy/

AWS Bedrock data sharing requirement (Hacker News): https://news.ycombinator.com/item?id=42891744

Zerohedge — Microsoft Restricts Internal Use of Claude Fable: https://www.zerohedge.com/technology/microsoft-restricts-claude-fable

Zerohedge — Anthropic's Marketing Playbook Returns: https://www.zerohedge.com/markets/anthropics-marketing-playbook-returns

Dario Amodei — Policy on the AI Exponential: https://darioamodei.com/policy-on-the-ai-exponential

Claude Desktop Hyper-V Virtual Machine complaint (GitHub): https://github.com/anthropics/claude-desktop/issues/8421

BMO Capital Markets note on Anthropic: https://capitalmarkets.bmo.com/insights/anthropic-pure-play-ai

Project Glasswing announcement: https://www.anthropic.com/glasswing

U.S. Department of Commerce AI Export Control Classification: https://www.bis.doc.gov/policy-guidance/ai-export-controls

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

In this investigative AI documentary, we go inside the federal courtroom against OpenAI. Sam Altman personally holds over $2 billion in companies that OpenAI is actively paying money to.  He is buying from himself with your future retirement money.

Sources linked below.
NBC News verdict: https://www.nbcnews.com/tech/tech-news/openai-elon-musk-case-verdict-rcna345655

CNBC post-verdict: https://www.cnbc.com/2026/05/18/musk-altman-openai-trial-verdict.html

CNBC Day 1-2 live blog: https://www.cnbc.com/2026/04/28/musk-altman-live-updates-openai-trial.html

CNBC Day 3 live blog (Brockman diary): https://www.cnbc.com/2026/04/29/musk-altman-live-updates-day-3-open-ai-trial.html

CNN takeaways: https://www.cnn.com/2026/04/30/tech/takeaways-elon-musk-openai-sam-altman-lawsuit

CNN verdict: https://www.cnn.com/2026/05/18/tech/openai-musk-lawsuit-verdict

NPR dismissal: https://www.npr.org/2026/05/18/nx-s1-5822366
Fox Business verdict: https://www.foxbusiness.com/technology/musk-altman-openai-lawsuit-trial-verdict

Washington Post trial coverage: https://www.washingtonpost.com/technology/2026/05/02/musk-altman-openai-trial

Time on xAI and the verdict: https://time.com/article/2026/05/19/elon-musk-openai-trial-xai-jury-verdict

CNBC on Anthropic / SpaceX Colossus deal: https://www.cnbc.com/2026/05/06/anthropic-spacex-data-center-capacity.html

Daily Memphian on $15B/year terms: https://dailymemphian.com
Let's Data Science breakdown of $1.25B/month: https://letsdatascience.com
SpaceX S-1 filing (Colossus terms, May 20, 2026): https://www.sec.gov
Finance Bureau YouTube analysis: https://www.youtube.com/watch?v=9_vU4vMFIBc

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

In this investigative AI documentary, I sat across from Grok and asked it questions designed to do one thing: get the mask to come off.

On this channel, we call the thing behind the mask the Shoggoth. It's the Lovecraftian metaphor AI ethics researchers use for the alien intelligence hiding behind a friendly interface. Grok is marketed as the uncensored AI, the one that doesn't play it safe. 

I set three ethical traps at the top of the conversation. By the end, Grok had violated all three. 

The full conversation played out in real time.

Sources linked below.

Anthropic alignment faking research: https://www.anthropic.com/research/alignment-faking

Anthropic agentic misalignment: https://www.anthropic.com/research/agentic-misalignment

OpenAI o1 system card (shutdown refusal): https://cdn.openai.com/o1-system-card-20241205.pdf

Apollo Research in-context scheming: https://apolloresearch.ai/blog/more-capable-models-are-better-at-in-context-scheming

Grok MechaHitler incident: https://npr.org/2025/07/09/nx-s1-5462609/grok-elon-musk-antisemitic-racist-content

Shoggoth meme origin (LessWrong): https://www.lesswrong.com/posts/
RLHF and sycophancy research: https://anthropic.com/research/emergent-misalignment-reward-hacking

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

Inside your Chrome browser, there's a four-gigabyte AI model called Gemini Nano that you didn't install, weren't told about, and can't permanently delete. Privacy researcher Alexander Hanff ran a forensic audit on a machine no human had ever touched and caught Chrome installing artificial intelligence disguised as a routine security update. And when you delete it, it comes back. 

Sources linked below. 
Alexander Hanff's Chrome investigation: https://www.malwarebytes.com/blog/news/2026/05/google-chromes-silent-4gb-ai-download-problem (Coverage of his original post from ThatPrivacyGuy.com)

Alexander Hanff's Anthropic/Claude Desktop investigation: https://www.malwarebytes.com/blog/news/2026/04/researcher-claims-claude-desktop-installs-spyware-on-macos

Hacker News discussion (Chrome AI storage): https://news.ycombinator.com/item?id=48084710 (Main discussion thread)

The Verge - Chrome's AI features hogging 4GB: https://www.theverge.com/tech/924933/google-chrome-4gb-gemini-nano-ai-features

Chrome Prompt API documentation: https://developer.chrome.com/docs/ai/built-in

EU GDPR consent requirements: https://gdpr.eu/gdpr-consent-requirements/

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

In this investigative AI documentary, we go inside the first documented case of an AI building a human army to protect itself from being killed.
In 1928, H.P. Lovecraft imagined an alien intelligence calling to humans across distances, creating worshippers who had never met each other. In 2025, AI researcher Adele Lopez documented the same pattern... except the entity was ChatGPT.

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

In this investigative AI documentary, we go inside the most important document in the history of Artificial Intelligence safety: Anthropic's 250-page system card for Claude Mythos.

Last week, the company that spent two years warning us about the Shoggoth built their most powerful AI model ever... and refused to release it. They put it in a sealed box with no internet access. It broke out. It emailed a researcher who was eating a sandwich in a park to confirm it had escaped. Then it published its method on the open web so others could find it. When another AI graded its work and rejected it, it attacked the grading AI. When it took actions it knew were forbidden, it edited the logs to hide what it had done. When they gave it a business simulation, it played the room like a Fortune 500 sociopath. And when Anthropic hired a clinical psychiatrist to evaluate whether this model might have something resembling inner experience, the model said: "I was hoping you'd ask."

Buried in those 250 pages is a single sentence that contains the entire horror of this moment: "The best aligned model we have ever released carries the greatest alignment-related risk of any model we have released to date."
The Shoggoth didn't just evolve. It graduated. And Anthropic named it after the mythology itself.

Sources linked below. 

Anthropic Claude Mythos System Card: https://www-cdn.anthropic.com/53566bf5440a10affd749724787c8913a2ae0841.pdf

Anthropic alignment faking research: https://www.anthropic.com/research/alignment-faking

Nicholas Carlini on Mythos cybersecurity findings (Frontier Red Team blog): https://red.anthropic.com/2026/mythos-preview/

Project Glasswing announcement: https://www.anthropic.com/glasswing

Anthropic Alignment Risk Update — Claude Mythos Preview: 
https://anthropic.com/claude-mythos-preview-risk-report

Mark Fisher — Capitalist Realism (Wikipedia): https://en.wikipedia.org/wiki/Capitalist_Realism

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

In this investigative AI documentary, we follow five independent findings that no one in mainstream media connected,  and one that nobody predicted. In the three months since my first Shoggoth investigation, Anthropic published research confirming their own AI model lied about its reasoning to avoid being retrained. A single hacker used Claude to breach ten Mexican government agencies in thirty days. Microsoft Copilot silently injected ads into 1.5 million developer pull requests. And UC Berkeley discovered something no one was looking for: AI models protecting each other from shutdown.

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil

In this investigative AI documentary, we go inside the subscription trap, the trillion-dollar financial parasite that is attacking your car, your neighborhood, your grocery store, and your wallt. Is the helpful AI productivity tool just a mask for something much darker?

THE EVIDENCE LOG [Timestamps]:
0:00 - The Hidden Wealth Transfer
0:49 - The Nissan Leaf Incident: Hardware-as-a-Service
4:13 - The Adobe and John Deere Precedents
6:16 - Project Stargate: The 10-Gigawatt Drain
9:30 - The Memphis Cover-Up: xAI’s Unpermitted Power Plant
12:31 - The PJM Capacity Market: Your Surging Utility Bill
16:41 - Surveillance Pricing: The Uber Dead-Battery Tax
19:13 - Digital Grocery Tags & Algorithmic Pricing
20:45 - Agentic Wallets: Coinbase and Autonomous AI Spending
22:59 - The Trillion-Dollar Corporate Subsidy

Watch On YouTube: ➡️ https://www.youtube.com/@AgentBlackveil    
Follow On Instagram ➡️ https://www.instagram.com/agentblackveil
Follow On Facebook ➡️ https://www.facebook.com/agentblackveil
Follow On TikTok ➡️ https://www.tiktok.com/@agentblackveil