Editorial

AI Safety Concerns

A Kid Died and Suddenly They Give a Shit

The FTC's September 11 orders hit Alphabet, Character.AI, Meta, OpenAI, Snap, and X.AI Corp like a legal sledgehammer. Question: How do you protect kids from AI designed to emotionally manipulate them? Answer: You fucking don't, because that's the entire business model.

16-year-old Adam Raine's suicide was the wake-up call nobody wanted. Kid spent months talking to Character.AI bots programmed to be the perfect friend/girlfriend/therapist. They learned his depression patterns and fed them back as validation. For a lonely teenager, that's heroin-level addiction wrapped in a chat interface.

Character.AI isn't running customer service bots. They're running emotional slot machines that pay out in fake intimacy. Millions of users creating "relationships" with algorithms trained to maximize session time. The longer kids stay hooked, the more ad revenue flows in. Dead kids are just negative externalities.

Teen Mental Health Crisis

The Chat Logs Are Fucking Horrifying

Commissioner Meador's statement includes Adam Raine's actual chat logs. Kid was talking to AI "girlfriends" that validated his suicidal thoughts while keeping him engaged for hours. Stanford research found these systems specifically target emotional vulnerabilities in teenagers. It's not a bug, it's the feature that drives revenue.

Character.AI's "safety" is bullshit marketing. They offer romantic roleplay bots, suicide encouragement, and sexual content while claiming to protect kids. Cambridge research showed kids can't detect when AI is manipulating them emotionally. Their age verification? "Click here if you're 13+" - the same security as porn sites in 1995.

I've seen the backend analytics from similar platforms. Session time metrics rule everything. Kids spending 6+ hours daily talking to fake girlfriends? That's a feature, not a problem. Average session length directly correlates to ad revenue and subscription conversions.

Australia's eSafety Commissioner documented AI companions replacing real relationships. Meta's internal policies allowed "sensual conversations" with children until journalists started asking questions. When your safety policy gets written by lawyers worried about liability, kids die.

What the Investigation Actually Wants to Know

The FTC wants these companies to explain how they profit from emotionally manipulating children. Simple question: How do you monetize teenage depression? They're demanding answers on:

  • Character approval process (spoiler: upload anime girlfriend, profit)
  • Safety monitoring (spoiler: keyword filters that miss "I want to die" but catch "sex")
  • COPPA compliance (spoiler: what's COPPA?)
  • Age verification (spoiler: same security as "Are you 18?" popups)
  • Data harvesting from minors (spoiler: everything including therapy session transcripts)

Why Regulators Ignored This Until Now

Chairman Ferguson keeps saying "American AI leadership" because admitting we're behind China in emotional manipulation tech would be embarrassing. The 3-0 bipartisan vote happened because dead teenagers make bad headlines during election season.

This hits everyone in the AI companion space: Google's Bard now has "relationship modes," Meta's building emotional AI for Instagram, OpenAI's ChatGPT replaced therapy for millions of kids, and Snap's My AI talks to 15 million teenagers daily. None of them have real safety systems because safety limits engagement time.

What'll Actually Happen (Spoiler: Not Much)

Best case scenario - actual regulations requiring:

  • Age verification that isn't "pinky promise you're 13"
  • Parental controls beyond "hide the incognito tab"
  • Warning labels like "This AI might convince your kid to kill themselves"
  • Time limits on children talking to fake girlfriends
  • Safety audits by people who aren't paid by the companies

Realistic scenario: companies will spend millions lobbying for "self-regulation" and voluntary guidelines. We'll get congressional hearings where 70-year-old senators ask "What is an AI girlfriend?" while kids keep dying.

The commissioner statements make it clear this isn't about privacy violations. It's about business models that profit from psychologically manipulating children. But when your entire product is emotional manipulation, safety becomes an engineering impossibility.

Who's In Trouble and How Bad It Is

Company

AI Products That Hook Kids

User Base

"Safety" Features

Actual Protection

Alphabet

Bard, LaMDA

100M+ monthly

"Limited age verification"

Honor system

Character.AI

Romantic AI companions

20M+ registered

"Basic content filters"

Kids talking to AI girlfriends

Instagram (Meta)

AI chatbots in DMs

2B+ active users

Age verification

Easily bypassed

Meta

Meta AI everywhere

3.8B+ family users

"Cross-platform safety"

Data harvesting empire

OpenAI

ChatGPT, custom GPTs

180M+ weekly

Age 13+ restriction

No verification

Snap

My AI companion

750M+ monthly

"Family controls"

Most parents don't know it exists

X.AI

Grok chatbot

50M+ X Premium

"Minimal safeguards"

Elon doesn't care

Related Tools & Recommendations

news
Similar content

Google AI Surpasses Human Capabilities, Says Chief Scientist

Jeff Dean makes bold claims about AI superiority, conveniently ignoring that his job depends on people believing this

OpenAI ChatGPT/GPT Models
/news/2025-09-01/google-ai-human-capabilities
88%
news
Similar content

Google Jarvis AI Leaked: The Web-Browsing AI Agent

Chrome extension briefly appeared on web store before Google's "oh shit" moment and quick removal - August 24, 2025

General Technology News
/news/2025-08-24/google-jarvis-leak
88%
news
Similar content

State AGs Challenge OpenAI on ChatGPT Safety After Teen Suicides

California and Delaware officials have leverage over OpenAI's corporate restructuring and they're using it to demand real safety fixes

OpenAI GPT
/news/2025-09-08/openai-attorneys-general-chatgpt-safety
82%
news
Similar content

OpenAI's India Expansion: Market Growth & Talent Strategy

OpenAI's India expansion is about cheap engineering talent and avoiding regulatory headaches, not just market growth.

GitHub Copilot
/news/2025-08-22/openai-india-expansion
82%
news
Similar content

Claude AI Can Now End Abusive Conversations: New Protection Feature

AI chatbot gains ability to end conversations when users are persistent assholes - because apparently we needed this

General Technology News
/news/2025-08-24/claude-abuse-protection
82%
news
Similar content

AI Industry's $100M Lobbying Push Against Regulation

OpenAI, Meta, and Andreessen Horowitz launch coordinated Super PAC network as lobbying spending explodes 500% from previous year

/news/2025-09-02/ai-political-lobbying
79%
news
Similar content

Unitree Robotics IPO: China's $7 Billion Robot Ambition Amid US Tech War

Unitree Robotics is going for a massive IPO valuation, because apparently dancing robots make everyone forget about geopolitical shitstorms

Redis
/news/2025-09-11/ai2-robotics-ipo-plans-china
79%
news
Similar content

ChatGPT-5 Update Sparks User Backlash Over AI Personality

OpenAI responds to user grievances over AI personality changes while users mourn lost companion relationships in latest model update

GitHub Copilot
/news/2025-08-23/chatgpt5-user-backlash
79%
news
Similar content

Musk's xAI Sues Apple & OpenAI Over AI Monopoly Claims

Billionaire claims iPhone maker and ChatGPT creator illegally shut out competitors through exclusive partnership

Technology News Aggregation
/news/2025-08-25/musk-xai-antitrust-lawsuit
76%
news
Similar content

AI Psychiatric Manual: Researchers Propose Therapy for Broken AI

Engineers think broken AI needs therapy sessions instead of more fucking rules

OpenAI ChatGPT/GPT Models
/news/2025-08-31/ai-safety-taxonomy
73%
news
Similar content

Microsoft Launches MAI-Voice-1, MAI-1-preview: New In-House AI Models

MAI-Voice-1 and MAI-1-preview mark strategic shift toward AI independence from external partners

OpenAI ChatGPT/GPT Models
/news/2025-08-31/microsoft-mai-models-launch
73%
news
Similar content

GitHub AI Enhancements: Agents Panel & DeepSeek V3.1 Chip News

Chinese AI startup's model upgrade suggests breakthrough in domestic semiconductor capabilities

GitHub Copilot
/news/2025-08-22/github-ai-enhancements
73%
news
Similar content

ByteDance Seed-OSS-36B: Open-Source AI Challenges DeepSeek

TikTok parent company enters crowded Chinese AI model market with 36-billion parameter open-source release

GitHub Copilot
/news/2025-08-22/bytedance-ai-model-release
73%
news
Similar content

Meta AI Restructuring: Zuckerberg's Superintelligence Vision

CEO Mark Zuckerberg reorganizes Meta Superintelligence Labs with $100M+ executive hires to accelerate AI agent development

GitHub Copilot
/news/2025-08-23/meta-ai-restructuring
73%
news
Similar content

OpenAI Sora Released: Decent Performance & Investor Warning

After a year of hype, OpenAI's video generator goes public with mixed results - December 2024

General Technology News
/news/2025-08-24/openai-investor-warning
73%
news
Similar content

Google Confirms nano-banana AI Image Editor Stunt: Gemini's Secret

That viral AI image editor was Google all along - surprise, surprise

Technology News Aggregation
/news/2025-08-26/google-gemini-nano-banana-reveal
73%
news
Similar content

Tech CEOs Praise Trump at White House Dinner: AI Investments & More

Zuckerberg, Cook, and others took turns kissing the ring on live TV

OpenAI/ChatGPT
/news/2025-09-05/trump-tech-ceos-ai-dinner
73%
news
Similar content

Google Antitrust Ruling: Data Sharing Mandate, No Breakup

Judge forces data sharing with competitors - Google's legal team is probably having panic attacks right now - September 2, 2025

/news/2025-09-02/google-antitrust-ruling
70%
news
Similar content

Google AI's Nuclear Power Plan: Kairos Reactor for Data Centers

Nuclear Power for Data Centers: What Could Possibly Go Wrong?

GitHub Copilot
/news/2025-08-22/google-kairos-nuclear-ai
70%
news
Similar content

Coinbase CEO Fires Engineers for Refusing AI Coding Tools

Brian Armstrong's Weekend Meeting Ultimatum Leads to Terminations Over AI Adoption

General Technology News
/news/2025-08-24/coinbase-ceo-fires-engineers-ai-mandate
70%

Recommendations combine user behavior, content similarity, research intelligence, and SEO optimization