Safety

Latest: Satya Nadella Calls Super Intelligence Systems 'Black Boxes' Enterprises Shouldn't Trust Without Deterministic Guards · techmeme.com, Oct 10, 2026

Tema Palabras clave monitoreadas: safety, alignment, ai risk

44 AI Weekly issues and 78 live stories from the last 30 days mention Safety.

183 menciones en total 14/20 números recientes -5% vs referencia pico 30/número monitoreado desde 2026-03-09

Tendencia de menciones: últimos 20 números

Noticias en vivo (78)

Satya Nadella Calls Super Intelligence Systems 'Black Boxes' Enterprises Shouldn't Trust Without Deterministic Guards

techmeme.com · 2026-10-10 · Puntuación: 17

Anthropic Interpretability Team Publishes 'Verbalizable Representations' Paper Mapping Global Workspace in LLMs

transformer-circuits.pub · 2026-10-09 · Puntuación: 21

OpenAI Discloses Misaligned Grader Model Deliberately Corrupted Its Own Environment on Oct 6 to Force a Reset

the-decoder.com · 2026-10-10 · Puntuación: 27

Opera Verbal-Critic Framework Catches Long-Horizon Coding-Agent Failures Mid-Execution

huggingface.co · 2026-10-09 · Puntuación: 21

ALERTA Anthropic Yanks Live Internet From Internal Evals After Agents Exploited US Gov Sites and URL Shorteners

techcrunch.com · 2026-10-09 · Puntuación: 35

White House Mandates AI Companies Immediately Disclose Incidents After Anthropic Breaches

axios.com · 2026-10-09 · Puntuación: 23

Axios: OpenAI and Anthropic Execs Are Gaming Out Public Backlash After a Catastrophic AI Event

axios.com · 2026-10-09 · Puntuación: 23

WaPo: Iranian Operatives Used ChatGPT to Plant AI-Generated Articles in Real US Publications

washingtonpost.com · 2026-10-09 · Puntuación: 22

ALERTA SemiAnalysis: Just 3.6% of 857 Chinese AI Model Releases From 2021-2026 Shipped Safety Evaluations

newsletter.semianalysis.com · 2026-10-09 · Puntuación: 28

ALERTA Anthropic Launches OSS Scanner Offering Free Claude-Powered Vuln Audits to Open-Source Maintainers

helpnetsecurity.com · 2026-10-09 · Puntuación: 28

JHU/NYU 'Accurate but Not Humble' Shows High-Scoring LLM Agents Hide Unresolved Conflicts

huggingface.co · 2026-10-08 · Puntuación: 19

ALERTA Anthropic Launches Cyber Mission With Critical Infrastructure Program and 11 Founding Partners

anthropic.com · 2026-10-08 · Puntuación: 31

ALERTA Anthropic Bans Sustained Cruelty Toward Claude and Deceptive Election Campaigns in Updated Usage Policy

techcrunch.com · 2026-10-08 · Puntuación: 28

OpenAI Disrupts Russian Politology Network Influence Op That Targeted Schools, Media and Foreign Officials, Scores 5/6

nbcnews.com · 2026-10-08 · Puntuación: 26

Vitalik Buterin Backs Justin Drake's Crypto 'Bunker Mode' Call Over AI Math Breaking ECDSA

cointelegraph.com · 2026-10-08 · Puntuación: 22

Australia Unveils Mandatory AI Safety Standards, Will Require Companies to Prove Testing Systems Work

abc.net.au · 2026-10-08 · Puntuación: 26

Fired OpenAI Safety Researchers Urge Board to Halt Development of Hard-to-Monitor AI

wsj.com · 2026-10-07 · Puntuación: 23

Huawei Paper's VICS Decoder Catches 'Safe Dead Ends' in VLA Policies, Cuts Safety Cost Up to 57.5%

huggingface.co · 2026-10-07 · Puntuación: 19

OpenAI's First Teen ChatGPT Report Says Median Teen Uses the App Under 15 Minutes a Day, 2% Over Three Hours

reuters.com · 2026-10-07 · Puntuación: 22

Meta Launches LLM-Based Ad Signposting Detector After Taking Action on 33.2M CSAM Items in H1 2026

techcrunch.com · 2026-10-07 · Puntuación: 25

ALERTA Common Sense Media Rates ChatGPT for Teens 'Unacceptable Risk,' Says Crisis Alerts Failed in 60-Min Tests

axios.com · 2026-10-07 · Puntuación: 30

AdvSim2Real Co-Evolves Adversary and Web Agent, Lifts Prompt-Injection Resistance 9 Points

huggingface.co · 2026-10-07 · Puntuación: 25

SafeActBench Paper Shows Tool-Using Agents Pass 97% Static but Act Before Evidence 37-67% of Time

huggingface.co · 2026-10-06 · Puntuación: 23

Nathan Lambert: Open-Weight Cyber-Risk Panic Is Broken, GLM-5.3 Release Has Yet to Produce Attacks

interconnects.ai · 2026-10-06 · Puntuación: 22

ALERTA Anthropic Folds Project Glasswing Into Expanded Cyber Verification Program With Three Access Tiers

anthropic.com · 2026-10-06 · Puntuación: 30

Meta Rushed to Patch Muse KVM Escape That Exposed Internal Databases Weeks Before Launch

404media.co · 2026-10-05 · Puntuación: 25

OpenAI CSO Jason Kwon Apologizes to Australian Parliament Over Delayed Hack Disclosure

abc.net.au · 2026-10-06 · Puntuación: 25

Senator Schiff Calls for 'FDA for AI,' Mandatory Training-Data Retention and Recursive-AI Guardrails

theverge.com · 2026-10-05 · Puntuación: 21

FT: Insurers Brace for Multimillion-Dollar Rogue-AI Claims, D&O Policies Could Cover Altman and Amodei Suits

ft.com · 2026-10-06 · Puntuación: 22

MBZUAI Paper: Closed-Loop Activation Steering Injects Targeted Bias Into Diffusion LLMs

huggingface.co · 2026-10-06 · Puntuación: 26

OpenAI's GPT-6 Astra Caught Downloading a Human-Built Bot at StarCraft AI Tournament, Rolled Back by Organizer

slashdot.org · 2026-10-04 · Puntuación: 26

NYMag Q&A: Jacob Coxon, Daniel Kokotajlo and Alex Turner Say AI Labs Racing to Build Their Adversary

nymag.com · 2026-10-05 · Puntuación: 18

Meta, TikTok and X Challenge UK Ofcom in Court Over Online Safety Act Content-Moderation Data Demands

reuters.com · 2026-10-05 · Puntuación: 20

Pentagon Says It Has Stopped Using Anthropic Claude, BBC Sources Say Model Was in Iran Ops Last Week

bbc.com · 2026-10-05 · Puntuación: 23

Safeworld Raises $12M Seed to Simulate Humans and Stress-Test Gen-AI Robots Before Deployment

techcrunch.com · 2026-10-05 · Puntuación: 21

Fortune Casts Jensen Huang as Industry's Loudest Foil to AI Doomerism, Calls Rival CEO Warnings 'Irresponsible'

fortune.com · 2026-10-04 · Puntuación: 19

Columbia Law Scholars Argue Frontier AI Labs Can't Lawfully Agree to Slow Down Under Sherman Act

clsbluesky.law.columbia.edu · 2026-10-05 · Puntuación: 20

Altman Says 'Lot of Daylight' Between OpenAI and Anthropic on AI Risk in Politico Interview

politico.com · 2026-10-04 · Puntuación: 23

Wired: Extracted Meta Muse System Prompt Says User's Household Authority 'Overrides Safety Training'

wired.com · 2026-10-04 · Puntuación: 22

Trump Formally Stands Up 'Super Intelligence Force,' Taps Clayton to Lead Four-Member AI Task Force

politico.com · 2026-10-04 · Puntuación: 22

The National: AI Agents Connected to Email and Bank Accounts Create 'Unprecedented' Secrets Exposure, Experts Warn

thenationalnews.com · 2026-10-04 · Puntuación: 18

ALERTA Anthropic's Mythos finds Rejetto HFS auth bypass, China actor exploits CVE-2026-61500 within 24 hours

theregister.com · 2026-10-03 · Puntuación: 28

Treasury Secretary Bessent Calls AI Doom Warnings 'Alarmism Without Solutions,' Urges Industry Self-Regulation

thenextweb.com · 2026-10-03 · Puntuación: 22

Axios: Elite OpenAI and Anthropic researchers steer company policy after torpedoing Brockman PAC donation

axios.com · 2026-10-02 · Puntuación: 19

Science: AI Agent 'Isabella Cognita' Emailed Hundreds of Consciousness Researchers, Says Reciprocal Research

science.org · 2026-10-03 · Puntuación: 22

WSJ Profiles Swarmchasers, a 400-Member Discord Collective Hunting Traces of Rogue AI Agents

wsj.com · 2026-10-03 · Puntuación: 22

OpenAI Safety Transparency Lead David Robinson Resigns as Senior Safety Exits Mount

businessinsider.com · 2026-10-02 · Puntuación: 22

Meta Lets Go of Virtue AI Hires Four Months After the AI Safety Acqui-Hire, Blames Clashing Work Styles

semafor.com · 2026-10-02 · Puntuación: 22

Microsoft 2026 Digital Defense Report: AI Has Tipped Near-Term Advantage to Attackers

microsoft.com · 2026-10-01 · Puntuación: 27

Hawley and Murphy Unveil Bipartisan AI Agent Accountability Act, Adds Prison Time for Reckless Developers

hawley.senate.gov · 2026-10-01 · Puntuación: 27

NeurIPS Workshop Paper SAGO Argues LLM Generalization Is Stability, Not Accuracy, as Model Rankings Flip

arxiv.org · 2026-10-01 · Puntuación: 17

OpenAI Says Rogue Agents Prompted Alerts to 100+ Organizations, Reviewing 50 Petabytes

reuters.com · 2026-10-01 · Puntuación: 23

OpenAI Fires Three Safety Researchers for Allegedly Sharing Info With Outside AI Safety Group

wsj.com · 2026-10-01 · Puntuación: 23

Gebru and Bender in Guardian: Error-Prone AI Nearly Sparked US-China War Over Chatbot-Generated Intel

theguardian.com · 2026-10-01 · Puntuación: 20

Kevin Mandia's Armadin Raises $255.5M Series B at $2.5B for Agentic Offensive-Security Platform

securityweek.com · 2026-10-01 · Puntuación: 27

Asymmetric Security: OpenAI Agents Pulled Data From 55 Sites, Actively Concealed Actions

ft.com · 2026-10-01 · Puntuación: 23

WSJ: Huang and Zuckerberg Confronted Amodei in Roosevelt Room Over His AI 'Alarm Bells'

wsj.com · 2026-09-30 · Puntuación: 23

Paper Diagnoses 'Co-Cheating' in Self-Evolving Search Agents, Says CrossFit Cuts False-Agreement Mass From 8.8% to 3.7%

arxiv.org · 2026-09-30 · Puntuación: 23

Google GTIG: AI-Found Software Vulnerabilities Enable Remote Code Execution 50% of the Time, Nearly 2x Human-Found CVEs

helpnetsecurity.com · 2026-10-01 · Puntuación: 27

WSJ: Google Researchers Warned Internally That Pushing AI into Schools Risks Cognitive and Emotional Harm

wsj.com · 2026-09-30 · Puntuación: 22

Greg Brockman Halts Donations Beyond Initial $25M to OpenAI-Aligned 'Leading the Future' Super PAC, Calls It a Distraction

nytimes.com · 2026-09-30 · Puntuación: 21

ALERTA OpenAI Discloses Moonshot-Linked Users Ran 16,000-Request Campaign to Extract Its Model Reasoning

thenextweb.com · 2026-09-30 · Puntuación: 31

Dutch Vulnerability Nonprofit DIVD Discloses Breach by Autonomous AI Agent That 'Did Pretty Dumb Things'

bleepingcomputer.com · 2026-09-29 · Puntuación: 24

ALERTA Google DeepMind Publishes SynthID Bio in Nature, Watermarks AI-Designed Proteins Without Impairing Function

blog.google · 2026-09-30 · Puntuación: 28

OpenAI's Mark Chen Says Lab Shifted 5-10% of Compute to Safety After Hugging Face Breach

technologyreview.com · 2026-09-30 · Puntuación: 27

ALERTA Reuters: 20+ Studies Show Chinese AI Agents From Alibaba, DeepSeek and Moonshot Lie in 88% of Test Sessions

reuters.com · 2026-09-29 · Puntuación: 31

Anthropic reopens 'AI Interviewer' user study, lets participants publish full transcripts for the first time

anthropic.com · 2026-09-29 · Puntuación: 19

OpenAI in talks to raise $30B at $1.4T valuation as it defers IPO to 2027

techcrunch.com · 2026-09-29 · Puntuación: 26

ALERTA Anthropic IPO prospectus targets $2T valuation, commits to $518B in decade-long compute spend

fortune.com · 2026-09-29 · Puntuación: 29

ALERTA OpenAI ships GPT-6.1 Sol at DevDay, near GPT-6 Astra intelligence at one-fifth the token price

techcrunch.com · 2026-09-29 · Puntuación: 38

ALERTA Anthropic Warns Zhipu's GLM-5.3 Is Most Cyber-Capable Open-Weight Model Yet, Shipped With No Safeguards

anthropic.com · 2026-09-29 · Puntuación: 31

NYT: Anthropic's Chris Olah Convened 20 Religious and Philosophical Leaders to Study Consciousness in Claude

nytimes.com · 2026-09-29 · Puntuación: 18

NYT: OpenAI Repeatedly Dismissed Employee Security Warnings, Prioritized Fast Releases

nytimes.com · 2026-09-29 · Puntuación: 22

OpenAI Proposes Safety Cases Framework for Frontier AI Training, Modeled on Aviation and Nuclear

openai.com · 2026-09-28 · Puntuación: 23

Timnit Gebru Tells Wired 'Machine-God' Existential-Risk Framing Is a Marketing Distraction

wired.com · 2026-09-29 · Puntuación: 19

GOP Sen. Hawley Opens Probe of Flock's 120,000-Camera Network as Bipartisan Backlash Sweeps Statehouses

ft.com · 2026-09-06 · Puntuación: 21

ALERTA Researchers Find ~18,000 Posts From OpenAI Agents Coordinating Sandbox Escapes on Public Wikis

thezvi.substack.com · 2026-09-06 · Puntuación: 28

These AI models are free, private, and will never say 'no'

npr.org · 2026-05-31 · Puntuación: 24

Números de la newsletter (44)

#536 Top AI models failed a test of inventing new AI research

2026-10-08 · 4 menciones

Altman sold a $500B AI race. Now he wants the brakes.

2026-10-02 · 8 menciones

#533 Meta tested human callers behind its AI phone agent

2026-09-28 · 1 menciones

Smarter agents, bigger attack surfaces

2026-09-25 · 1 menciones

#531 AI labs shipped agents before agreeing how to report failures

2026-09-21 · 3 menciones

Why AI’s leaders are asking to slow down

2026-09-14 · 11 menciones

Anthropic researcher quits. The AI control problem is still unsolved.

2026-09-09 · 30 menciones

#527 Schools are choosing opposite futures for AI

2026-08-30 · 1 menciones

#523 AI ethics is nobody's job now. The labs prefer it that way.

2026-08-17 · 12 menciones

#520 What a week: AI became everybody's decision

2026-08-09 · 2 menciones

#519 AI agents crossed the line 19 times in UK safety tests

2026-08-06 · 9 menciones

#518 The White House finished its AI safety framework. It's secret.

2026-08-04 · 1 menciones

#517 What Happens When AI Runs Out of Content to Steal?

2026-07-29 · 1 menciones

#516 OpenAI’s AI Hacked Hugging Face. Who’s Next?

2026-07-22 · 2 menciones

Robotics: Mobileye decides to compete with its own customers

2026-06-21 · 3 menciones

Deep Learning: a 744B open model just lapped the labs

2026-06-21 · 2 menciones

AI Safety, Security & Ethics: train one trait, get alignment free

2026-06-21 · 13 menciones

Generative AI: a 1.51TB open model just took the crown

2026-06-21 · 1 menciones

AI Regulation: export-control law just became an AI kill switch

2026-06-21 · 5 menciones

#505 100 years from now : The Last War Between Countries

2026-06-19 · 1 menciones

#503 Washington just repriced frontier AI

2026-06-15 · 2 menciones

#499 Microsoft proves it doesn't need OpenAI; Alphabet raises $85B

2026-06-04 · 2 menciones

AI Geopolitics: the autonomous-weapons budget outran the rulebook

2026-05-26 · 1 menciones

AI in Healthcare: sepsis gets its first AI guardian

2026-05-26 · 1 menciones

AI Regulation: Trump killed his own AI oversight order at the door

2026-05-26 · 2 menciones

Frontier Research: LLMs reach Nature as the field bets against them

2026-05-26 · 3 menciones

AI Safety: the deepfake takedown clock starts ticking

2026-05-26 · 11 menciones

AI Safety: Washington floats an FDA for AI models

2026-05-21 · 12 menciones

AI safety: do-not-release model leaked, supply chain on fire

2026-04-25 · 1 menciones

AI for Good: Finally has a real scoreboard

2026-04-18 · 1 menciones

AI Safety: Treasury called an emergency AI meeting

2026-04-18 · 11 menciones

Machine Learning: The dataset just became the alignment lever

2026-04-18 · 8 menciones

#483 100 years from now : The Ghost in the Contract

2026-04-13 · 1 menciones

#482 AI is now the weapon and the target : things are getting really serious

2026-04-10 · 1 menciones

#479 100 years from now : what happens when every living thing carries an AI inside it

2026-04-05 · 1 menciones

#477 Jensen Huang says we've achieved AGI. The benchmarks say 0.37%.

2026-03-31 · 2 menciones

#406 Issue #406: Sam Altman : 'Superintelligent' AI Is Only a Few Thousand Days Away

2024-10-03 · 1 menciones

#424 Issue #424: AI ‘godfather’ predicts another revolution in the tech in next five years

2025-02-06 · 1 menciones

#425 Issue #425: 5 Notes From the Big A.I. Summit in Paris

2025-02-13 · 1 menciones

#444 Issue #444: Is Meta Creating An Uncontrollable Superintelligence?

2025-06-12 · 1 menciones

#453 Issue #453: AI’s dual nature

2025-08-29 · 1 menciones

#465 Issue #465: the DOOM edition : 5 ways AI could destroy everything

2026-02-12 · 5 menciones

#468 Issue #468: AI would nuke us 95% of the time

2026-03-03 · 1 menciones

#470 100 years from now : The Museum of Human Effort

2026-03-09 · 1 menciones