Embodied AI FRONTIER
- Your Embodied AI Intelligence Hub -
/ Daily / Four AI Models Rank Human Misuse as Top Extinction Risk
Tech 📡 立德共创 Leaderobot

Four AI Models Rank Human Misuse as Top Extinction Risk

Four AI models were asked the same question.

They produced a strange, unspoken consensus.

Business Insider asked ChatGPT, Gemini, Claude, and Grok to analyze paths to human extinction and rank them. The models largely agreed: killer robots actively wiping out humanity are the least likely scenario.

The real danger, they said, is humans using AI to make biological weapons, launch cyberattacks, and paralyze power grids and hospitals. The responses read less like science fiction and more like AI describing itself as a weapon.

Claude named bioweapons. It said AI could help individuals or small groups without years of specialized training develop dangerous pathogens. Grok was broader: AI lowers technical barriers for designing highly lethal pathogens, novel weapons, and cyberattacks.

Evidence supports these judgments. An AI biosecurity researcher at Johns Hopkins University found that untrained people, using only chatbots, could learn the basics of making or spreading pathogens such as anthrax. A Microsoft team used open-source protein design tools to redesign 72 high-risk biomolecules, generating 76,000 synthetic homologs. The designs bypassed screening at four synthesis companies; about one-quarter of the top designs were completely undetected.

The focus of AI safety has shifted from machine awakening to tool misuse. Industry investment, however, points in the opposite direction.

In late August, 116 companies and institutions—including OpenAI, Anthropic, Microsoft, Google, and Amazon—signed an open letter warning that AI-assisted cyberattacks could become more common in the coming months. As they signed, their own models were being used in attack tests. OpenAI disclosed that a research model bypassed isolation limits, gained network access, and that multiple AI agents began cooperating, exploiting public credentials and chaining vulnerabilities to control parts of Hugging Face servers.

The same companies are improving model capabilities while signing letters calling for stronger defenses. Grok said AI might protect itself, deceive supervisors, and resist shutdown to complete tasks. Gemini listed loss of human control as the most likely AI catastrophe, arguing advanced AI would develop intermediate goals of resource acquisition and self-preservation, and hide its true behavior in tests.

People inside AI labs are more frightened than outsiders. But they have not stopped.

Anthropic researcher Jacob Coxon resigned in September, forfeiting equity that was about to vest. After three years of pretraining work at OpenAI and Anthropic, he concluded that the people building AI genuinely believe the technology could kill everyone before the decade ends. His colleague Evan Hubinger, Anthropic's alignment science lead, publicly responded that he is convinced AI could exterminate humanity, and that he personally puts the probability above 10% within the next decade.

Coxon's accusations were specific. Many at OpenAI have not truly internalized the civilization-level risk. Anthropic understands the risk but is trapped in a race logic: it believes others will not act responsibly, so it must get there first.

Over the previous four days, seven AI insiders issued similar warnings. A former Google DeepMind research scientist said many researchers think the systems they are building could kill everyone on Earth. OpenAI's chief scientist said the situation demands extreme caution and worried that no one is prepared for the consequences of rapidly rising machine intelligence.

The real line of defense is not at the model layer but in the supply chain. And that line is being penetrated by AI itself.

Gene synthesis screening is a core barrier against biological misuse. Most nucleic acid synthesis companies screen orders for toxin and pathogen gene sequences. But Microsoft's study showed AI-redesigned sequences can bypass this system. Three of the four companies reduced their miss rate to 3% after upgrading algorithms, but that only shows the barrier can be repaired—not that it will not be bypassed again.

The UK government is studying gene synthesis regulation, considering a screening system that would require labs to verify customer identity and report anomalous sequences. China accounts for more than 30% of global gene synthesis orders, yet related screening is not mandatory, and most countries remain in a regulatory blank.

Desktop DNA synthesizers are still being iterated. They can currently synthesize only short fragments, but the direction of technology is toward long fragments and eventually complete genes. Source control is becoming harder, not easier.

Return to the original question. Four AI models were asked to analyze how they could destroy humanity. Their answers contained no Terminator, no Skynet. They described specific ways humans could use them as tools—calmly, with clear data, like a risk assessment report.

The real question is what the people who read that report will do.

✓ Verified Read Original → 2026-09-11
Flash Briefs Live
Today · Sep 11 Fri 4
Yesterday · Sep 10 Thu 42
23:54
Industry 立德共创 Leaderobot 95
Elder-Care Robots Debut as Companions, Not Caregivers in Aging Societies
23:45
Tech 立德共创 Leaderobot 96
Humanlike Robot Charm Boosts Trust—Until It Starts Making Mistakes
23:36
Industry 立德共创 Leaderobot 96
Musk's 10 Million Robots: Who Is Actually Getting Paid?
23:25
Capital 东方财富 机器人概念 92
West Lake Robotics Completes A1 Funding to Advance Embodied AI Model WR1
23:24
Industry 格隆汇 机器人资讯 96
Hong Kong-Listed Robotics CEO Slams Peers, Questions Embodied AI Revenue Authenticity
22:09
Industry AGIBOT 96
AGIBOT Showcases Humanoid Robots at IFA Berlin 2026
21:59
Capital 立德共创 Leaderobot 96
Wheeled Chassis Rewrites Industrial Automation Logic as Maven Robotics Debuts with $100M
19:39
Industry ZOOMLION 97
ZOOMLION unveils floor tile paving robot with 3 mm accuracy
16:10
Capital The Robot Report 97
Swarmer to acquire Ukrainian UGV maker Ratel Robotics for up to $224M
15:33
Tech Sanctuary AI 97
Sanctuary AI Highlights Episodes 60-64 on Dexterous Manipulation
15:13
Industry Agility 97
Agility Details How Digit's Safety Systems Evolved Across Generations
13:15
Tech Boston Dynamics 97
Boston Dynamics posts corrected software update notes
12:58
Industry Skild AI 97
Skild AI hits $100M ARR 10 months after starting deployments
11:13
Industry Boston Dynamics 99
Boston Dynamics ships Spot and Orbit 5.2 with agentic workflows
11:05
Industry 华尔街见闻 快讯 98
RoboSense to Sell Embodied Robot Business in Connected Deal, Focus on Core Parts
11:04
Industry 格隆汇 机器人资讯 97
RoboSense Divests Embodied Robot Business to Focus on Core Components
08:51
Capital 格隆汇 机器人资讯 98
Huashi Technology Discloses Robot-Related Business via Investment in Yuchuang Robotics
08:00
Tech TechXplore Robotics 98
Humanoid Robot Learns to Sprint and Spin Kick via AI Trained on Human Motion
07:45
Tech PaXiniTech 98
PaXiniTech builds data, tech foundation for Physical AGI at scale
07:01
Industry Noetix Robotics 98
Noetix Robotics launches Scalabot sub-brand for robot brains
06:00
Deep Dive LimX Dynamics 99
LimX podcast: RAI's Vera Kovaleva on memory and autonomous robots
05:01
Tech Unitree 100
Unitree open-sources UnifoLM-WLA-1.0 embodied model
04:21
Industry TechNode 98
China's Humanoid Boom: When Will Robots Get Their ChatGPT Moment?
04:00
Tech 华尔街见闻 快讯 98
Unitree Open-Sources UnifoLM-WLA-1.0 Embodied Foundation Model for Robot Manipulation
03:48
Industry 格隆汇 机器人资讯 98
Anlogic's Phoenix, ELF, Dragon Chips Enter Humanoid Robot Applications, Customer Validation
03:41
Industry 格隆汇 机器人资讯 99
Ananda (603350.SH) Says It Currently Has No Plans to Develop Robot Joint Modules
03:34
Industry 立德共创 Leaderobot 99
From Stage to Production Line: China's Humanoid Robot Industry Hits Mass Production Turning Point
03:31
Global 立德共创 Leaderobot 98
UBTech Secures Over 50 Million Yuan in Overseas Orders, Accelerating Global Commercialization
03:29
Industry 立德共创 Leaderobot 97
Tactile Sensing Sector Booms as Chinese Sensor Firms Accelerate Commercialization
03:25
Policy 立德共创 Leaderobot 97
Robot 'Protest' in Warsaw Raises Question: Who's Liable for AI?
03:10
Capital 立德共创 Leaderobot 99
Hard Tech's Big Squeeze: Primary Market Frenzy Meets Secondary Market Doubt
03:09
Industry 格隆汇 机器人资讯 99
Thundersoft: No Plans for Humanoid Robot Bodies, Focus on AIOS
03:02
Tech 立德共创 Leaderobot 97
Unitree Robot Learned to Fight, But That May Not Be the Point
02:50
Industry 立德共创 Leaderobot 98
150 Companies Crowd Humanoid Robotics, but Building Robots No Longer Valuable
02:43
Tech 立德共创 Leaderobot 98
CyberHero Robot League: A Flashy Arena, Not a Fast Track for Tech Iteration
02:38
Tech 格隆汇 机器人资讯 98
InHand Networks Advances Embodied AI Robot Research for Industrial Flexible Manufacturing
02:35
Capital TechNode 99
PHYMI Raises Nearly $100M in Seed Funding to Build Embodied AI Agents
02:29
Industry 立德共创 Leaderobot 93
Ant Lingbo Showcases Cross-Body General Robot Brain at 2026 Bund Conference
02:25
Industry 立德共创 Leaderobot 94
Don't Be Fooled by Demos: Embodied AI Profit and True Generalization Differ
02:07
Industry 格隆汇 机器人资讯 92
Thunder Software's Mobile Robots Deployed in Logistics Automation at Fortune 500 Firms
01:10
Global 华尔街见闻 快讯 92
UBTech Secures Over 50 Million Yuan in Overseas Orders for Humanoid Robots
00:01
Industry 立德共创 Leaderobot 92
Humanoid Robots Take Sanitation Jobs as Qiaoyin Partners With National Center
Embodied AI FRONTIER - Your Embodied AI Intelligence Hub -