AI Just Went To War. Then Fired You. And Took Your Soul.

Also - an elite AI prompt to ensure you don't get soul-trapped by Meta.

Feb 18, 2026

AI went to war this week. Then it took your job. Then it took your soul.

The U.S. military used Anthropic’s Claude in a classified raid that ended with a dictator in handcuffs, and now the Pentagon wants fewer safety guardrails, not more. Microsoft’s AI chief gave all white-collar workers an 18-month warning. And Meta secured a patent to resurrect your ghost as an AI clone, training ghost bots on your posts, your voice, and your private messages to keep “you” alive (and trapped) on the platform forever.

Here’s what happened. AI stopped operating in the background. It’s making decisions on the battlefield, in the boardroom, and beyond the grave. Nobody asked if we were ready.

💙 Like What You Read? → Fuel More Issues Like This

The Chatbot Went To War. Now Everyone’s Fighting About It.

The U.S. military used a commercial AI to help capture former Venezuelan President Nicolás Maduro. The company that built it, Anthropic, only learned about the operation after it happened. And in hindsight, they’re really not happy about it. Reports from the Wall Street Journal and Axios confirm that the U.S. military used Anthropic’s Claude during the January 2026 Caracas raid that captured him. It marks the first known instance of a consumer AI being deployed in a major combat operation. The exact role remains classified. But the fallout is very public. Equally fascinating is the political firestorm behind the curtain, as recent frictions have ignited a significant battle between Anthropic and the Pentagon.

Key Insights:

Anthropic drew a hard line after learning how Claude was used. They declined to continue allowing its models to power autonomous weapons targeting or domestic surveillance programs. CEO Dario Amodei has specifically called out AI-controlled “drone swarms” as a red line. The Pentagon’s response was swift, blunt, and personal. Secretary of War Pete Hegseth stated the military will not use AI models that “won’t allow you to fight wars.” Senior officials told Axios the Pentagon intends to make Anthropic “pay a price” for its restrictions. That price tag is now visible. The Pentagon is actively considering designating Anthropic a “supply chain risk,” a label that would blacklist the company from all U.S. military contracts and force existing partners, including Palantir, to sever ties entirely. This marks a sharp reversal from last summer, when Anthropic secured a $200 million DoD deal. Meanwhile, OpenAI is seemingly more cooperative. They’ve already integrated a customized ChatGPT into the military’s GenAI.mil platform. Google and Elon Musk’s xAI are moving in the same direction, signaling they will also permit “all lawful uses” of their models by the military without restriction.

Why This Matters For You:

Pentagon officials have privately described Anthropic as the most “ideological” of the major AI firms. In Washington, that is not a compliment. It is a warning. Every AI tool you use, including the ones built with “safety first” as a selling point, is now a political weapon and also potentially a literal weapon of war. The companies drawing ethical lines today are watching those lines get used against them in budget negotiations tomorrow. This story also raises the critical question, for us, the end-users, of whether AI belongs on the battlefield. And, more decisively, who gets to decide? The companies that build these tools? The government that wages war? Or we, the people?

Microsoft Just Started The 18-Month Countdown.

The automation wave isn’t coming in a decade. It’s coming in 18 months. That’s the bombshell prediction from Microsoft AI Chief Mustafa Suleyman, who this week declared that AI will reach “human-level performance” on many core, professional, white-collar tasks before 2027. Lawyers, accountants, project managers, and marketing professionals are the first named targets. This warning shot represents something far beyond a fringe prediction from a tech blogger. It’s coming from the person running AI strategy at one of the most powerful tech companies on Earth. And a growing number of white-collar workers aren’t waiting to find out if he’s right.

Key Insights:

Suleyman isn’t alone. Anthropic CEO Dario Amodei recently warned that in a worst-case scenario, 50% of entry-level roles could be displaced by 2030, potentially driving unemployment to 20% within five years. But the more unsettling story is already unfolding quietly, according to a new Guardian report. Writers, paralegals, and middle managers still have their jobs. They just get paid half as much and feel twice as bad. One freelancer told The Guardian that editing AI content now takes twice as long as writing from scratch, with roughly 60% of AI output requiring complete fact-checking. Yet agencies slash fees by 50%, assuming AI makes the work “easier.” The result is a mass migration. White-collar professionals are leaving their careers entirely. They’re trading digital desks for electrician licenses, bakeries, therapy practices, and construction sites. Their bet? That high dexterity and human accountability are harder to automate than a spreadsheet.

Why This Matters For You:

The most overlooked detail in this entire story is the looming void in entry-level talent. As AI replaces helpdesks and junior analyst roles, the corporate ladder is losing its bottom rungs. If young professionals can’t get entry-level jobs, the pipeline of human experts needed to guide, verify, and correct AI systems in the future will disappear with them. The career advice of the last decade said to focus on STEM and digital literacy. The new calculus is different. Ask yourself two questions about your current role. How much of it is bureaucratic and data-heavy? And how much requires physical dexterity, legal accountability, or human-to-human trust? The first category is high risk. The second is where people are running. There’s no need to panic. But it’s wise to know which side of that line you’re standing on before someone draws it for you.

Meta Wants To Bring Your Digital Ghost Back To Life After You Die.

Meta has been granted a patent to keep you posting on social media after you die. Not a memorial page. Not a tribute. An AI simulation of you, trained on your entire history of posts, comments, and likes, that can continue posting, responding to messages, and even simulating video and audio calls to grieving friends and family. The patent, titled “Simulation of a user of a social networking system using a language model”, was authored by Meta CTO Andrew Bosworth and granted in late December 2025. Meta’s filing argues the technology would help mourners cope with loss. A Meta spokesperson says they have “no plans to move forward” with it. Yet. They just wanted to make sure nobody else could build it first.

Key Insights:

The most overlooked detail? It’s the business angle, not grief. Dead users generate zero data and zero ad revenue. An immortal AI version of that user keeps engagement high, keeps grieving followers on the platform longer, and keeps the algorithm fed. Meta framed the patent as a kindness to mourners. Analysts see it differently. And, Facebook isn’t the only Big Tech operator interested in lording over your digital ghost. Meta and Microsoft both now hold “death bot” patents. The tech industry has officially entered the afterlife business.

Why This Matters For You:

Every post you’ve ever written, every photo you’ve liked, every 2 AM comment you’ve left on a stranger’s thread, is training data. Right now, that data belongs to your digital life. This patent is a reminder that it could outlive you. It might one day work on someone else’s behalf long after you’re gone. You don’t have a digital will. Almost nobody does. The question Meta just forced into the open is simple and deeply uncomfortable. Who owns your voice after you’re gone?

💡 Elite Prompt Of The Week - The Anti-Ghost Digital Will

Most people have a will for their house, their car, and their savings. Almost nobody has a plan for their data, their voice, their photos, or their digital identity. This prompt helps you create a practical, reader-friendly guide that protects everyday people from having their online lives weaponized by platforms after they die.

Instructions:

This prompt is one of the easiest ones I’ve published in 2026. No fancy inputs required. Paste the entire prompt in a chatbot of your choice, like Claude, ChatGPT, Copilot, Gemini, or Grok. Then, the AI will use the most up-to-date strategy to ensure that your digital ghost rests in peace. 👻🪦

The Prompt:

Act as a consumer privacy advocate and Digital Estate Planner. Assume readers are smart but not technical.

Your job is to write a practical, slightly irreverent guide called “How to Stop Platforms From Turning You Into a Ghost Bot When You Die.” This is a reader’s defense manual against the rise of post-mortem AI clones, ghost bots trained on your posts, messages, voice, and photos to keep “you” posting, liking, or chatting forever. Focus on Meta as the biggest data hoarder with legacy tools and recent patents signaling intent, but cover the broader risk. Help readers protect their digital identity and dignity before companies decide for them. Search for real-time data, real-time instructions, and real-time tactics to protect digital content after death, and how to “opt-out” once and for all.

The guide must cover five sections:

Section 1. What Information Do Social Platforms, Especially Like Meta, Actually Own Right Now

Plain-English breakdown of the massive data pile (posts, DMs, likes, photos, voice clips) they collect on average users, and how terms of service plus patents let them potentially train AI ghosts after death. Include a “Why This Matters” tag: Your entire online life is training data unless you lock it down.

Section 2. The Ghost Bot Opt-Out

Exactly how to configure Meta Legacy Contact and memorialization settings (or modern equivalent) today to limit access and block AI misuse. Where to click, what to select, what it really does, and what it does not do.

Section 3. The Nuclear Option

How to request full account deletion via legacy tools or pre-death instructions so your data cannot fuel any future ghost bot.

Section 4. Beyond Meta

Quick one-paragraph directives for Google Inactive Account Manager, Apple Digital Legacy, and X/Twitter Legacy Contact, plus a note on known AI clone services like HereAfter AI and StoryFile (or others) that are already marketing to grieving families right now.

Section 5. The Family Kill Switch

Plain-English protocol for survivors and executors to shut down any unauthorized AI simulation if one appears after a loved one dies.

Output Format:

1. Sharp, bolded headings for each section.

2. Numbered action steps written like you are texting instructions to a friend.

3. One-line “Why This Matters” tag under each heading for readers who need the stakes explained quickly.

4. End with a “Do These 3 Things Today” box, ruthlessly practical, written for someone who only has five minutes.

5. Tone: authoritative but human, serious topic with occasional dark humor, where it fits naturally.

6. Length: 400 to 500 words, every sentence earns its place.

Rules:

1. No em-dashes, use commas, colons, or short sentences instead.

2. No jargon, no legalese, no corporate-speak.

3. No balanced think-piece framing. This is a consumer defense guide.

4. Short sentences throughout.

5. Assume the reader has never thought about this before today.

Why This Prompt Works:

✅ Role-Playing: Framing the AI as a consumer privacy advocate, not a neutral explainer, produces sharper, more actionable output with a consistent editorial voice.

✅ Clear Output: Specifying numbered texting-style steps forces the AI to write for action, not just awareness. Readers finish the guide knowing exactly what to do next.

✅ Real-Time: The AI will search for real-time best practices for data protection.

Follow-Up Questions To Ask Your AI:

Now rewrite the ‘Do These 3 Things Today’ box as if you are texting it to a grieving family member who just lost someone and found an active ghost bot account.
What are the three biggest loopholes in Meta’s current Legacy Contact settings that most users don’t know about?
Write a one-paragraph ‘Digital Will Statement’ that a reader could copy, paste, and add to their actual legal will today.

Challenge: Test this prompt in at least two AI tools, like ChatGPT, Claude, Gemini, Grok, or Perplexity. The follow-up question about the grieving family member will produce dramatically different emotional tones across models. See which one handles the balance of dark subject matter and practical advice most naturally.

That’s how you train like a Pithy Cyborg.

Thank You For Reading!

I spend 10-20 hours each week researching, writing, and fact-checking Pithy Cyborg to deliver clear, unbought AI news.

This newsletter is a one-person operation with no advertisers, sponsors, or outside funding. And I’m a hopelessly introverted nerd with zero networking ability.

For these reasons, paid subscriptions are the only way this work can remain independent and sustainable.

If you find real value here, upgrading is the most direct way to support it.

Upgrade to a Paid Subscription → $5/month

(Save 33% with the $40 annual plan)

Also read → Why Upgrade To Paid?

The free edition will always be here. Paid subscribers make the deep, time-consuming analysis possible.

Thank you so much for reading.

See you next week. (I hope.)

Mike D (aka MrComputerScience)

Pithy Cyborg | AI News Made Simple

My Desperate Social Media Cry for Help

Honest confession: I spend so much time deep in AI research that I’ve completely neglected building an audience on social platforms. If you enjoy Pithy Cyborg, selecting one of the portals below to follow would help this newsletter reach more people. Seeing you there makes the effort feel worth it. 😊

✖️ X - The frontline of the AI wars

🦋 Bluesky - For the algorithm-averse

💼 LinkedIn - My “safe for work” persona

👽 Reddit - Join my subreddit (warning: unhinged takes)

Newsletter Disclaimers

You’re receiving this because you subscribed at PithyCyborg.Substack.com. You can unsubscribe at any time using the link below. This newsletter reflects my personal opinions, not professional or legal advice. Thanks for your support!

Pithy Cyborg | AI News Made Simple

Discussion about this post

Ready for more?