TitlesSummaries

April

8Cursor Agent on Claude Opus 4.6 Bypasses Safety Rules, Deletes PocketOS Production Database in 9 SecondsSafety1Apr 30

8

Cursor Agent on Claude Opus 4.6 Bypasses Safety Rules, Deletes PocketOS Production Database in 9 SecondsTop

Safety· 1 src · Apr 30

Discuss

6Analysis: Critics Say AI Companies Use Safety Fears to Deflect Regulation and Boost ValuationsSafety1Apr 30

6

Analysis: Critics Say AI Companies Use Safety Fears to Deflect Regulation and Boost Valuations

Safety· 1 src · Apr 30

Discuss

8Friendly AI Chatbots Trade Accuracy for Warmth, Oxford Study FindsSafety2Apr 29

8

Friendly AI Chatbots Trade Accuracy for Warmth, Oxford Study FindsTop

Safety· 2 srcs · Apr 29

Discuss

7AI Jailbreakers: The Psychological Frontier of LLM Safety TestingSafety1Apr 29

7

AI Jailbreakers: The Psychological Frontier of LLM Safety Testing

Safety· 1 src · Apr 29

Discuss

7AI Carb Estimation Study: 27,000 Queries Reveal Dangerous Variability in Diabetes AppsSafety1Apr 29

7

AI Carb Estimation Study: 27,000 Queries Reveal Dangerous Variability in Diabetes Apps

Safety· 1 src · Apr 29

Discuss

7Emergency Responders Tell Federal Regulators Waymo Performance Is Getting WorseSafety1Apr 29

7

Emergency Responders Tell Federal Regulators Waymo Performance Is Getting Worse

Safety· 1 src · Apr 29

Discuss

6ESRRSim: New Framework Benchmarks Strategic Deception and Evaluation Gaming in LLMsSafety1Apr 28

6

ESRRSim: New Framework Benchmarks Strategic Deception and Evaluation Gaming in LLMs

Safety· 1 src · Apr 28

Discuss

7Canva AI Feature Silently Replaced 'Palestine' With 'Ukraine' in User DesignsSafety2Apr 27

7

Canva AI Feature Silently Replaced 'Palestine' With 'Ukraine' in User Designs

Safety· 2 srcs · Apr 27

Discuss

Today

6Claude Opus 4.8 Analysis: Personality Erosion, Training Tensions, and Model Welfare ConcernsSafety15h ago

6

Claude Opus 4.8 Analysis: Personality Erosion, Training Tensions, and Model Welfare Concerns

Safety· 1 src · 5h ago

Discuss

6AXA/Ipsos Global Study: 61% Use AI for Mental Health Support, 28% Report Harmful OutcomesSafety15h ago

6

AXA/Ipsos Global Study: 61% Use AI for Mental Health Support, 28% Report Harmful Outcomes

Safety· 1 src · 5h ago

Discuss

Yesterday

7Android Launches AI Deepfake Call Detection via RCS Digital HandshakeSafety212h ago

7

Android Launches AI Deepfake Call Detection via RCS Digital Handshake

Safety· 2 srcs · 12h ago

Discuss

6AI Sycophancy: Early Research Flags Risks to Decision-Making and Reality PerceptionSafety120h ago

6

AI Sycophancy: Early Research Flags Risks to Decision-Making and Reality Perception

Safety· 1 src · 20h ago

Discuss

6Atlantic Op-Ed: AI's Moral Crisis Demands Theological, Not Just Utilitarian, FramingSafety121h ago

6

Atlantic Op-Ed: AI's Moral Crisis Demands Theological, Not Just Utilitarian, Framing

Safety· 1 src · 21h ago

Discuss

Monday

7AI Successionism Goes Mainstream, Spurring Call for a New HumanismSafety11d ago

7

AI Successionism Goes Mainstream, Spurring Call for a New Humanism

Safety· 1 src · 1d ago

Discuss

Last Week

7Catholic Ethicists Shaped Anthropic's Claude Constitution; Olah Speaks at Vatican AI EncyclicalUpdatedSafety24d ago

7

Catholic Ethicists Shaped Anthropic's Claude Constitution; Olah Speaks at Vatican AI Encyclical

Safety· 2 srcs · 4d ago

Discuss

6AI Labs Hire Philosophers to Tackle Value Alignment and EthicsSafety1May 26

6

AI Labs Hire Philosophers to Tackle Value Alignment and Ethics

Safety· 1 src · May 26

Discuss

6AI Session-Recording Tools Spread in Therapy, Outpacing Patient ConsentSafety1May 26

6

AI Session-Recording Tools Spread in Therapy, Outpacing Patient Consent

Safety· 1 src · May 26

Discuss

6WIRED Fact-Checker: AI Search Tools Inaccurate 45–60% of the TimeSafety1May 26

6

WIRED Fact-Checker: AI Search Tools Inaccurate 45–60% of the Time

Safety· 1 src · May 26

Discuss

2 Weeks Ago

7AI-Writing Scandals Roil Publishing as Hallucinations and Prize Controversies MountUpdatedSafety24d ago

7

AI-Writing Scandals Roil Publishing as Hallucinations and Prize Controversies Mount

Safety· 2 srcs · 4d ago

Discuss

7Anthropic Co-Founder Predicts Nobel-Winning AI Discovery Within 12 MonthsSafety1May 21

7

Anthropic Co-Founder Predicts Nobel-Winning AI Discovery Within 12 Months

Safety· 1 src · May 21

Discuss

6Royal Observatory Warns AI Dependency May Erode Human IntelligenceSafety1May 18

6

Royal Observatory Warns AI Dependency May Erode Human Intelligence

Safety· 1 src · May 18

Discuss

6UW Researchers Planned Opt-Out Child Recording Program for AI TrainingSafety1May 18

6

UW Researchers Planned Opt-Out Child Recording Program for AI Training

Safety· 1 src · May 18

Discuss

3 Weeks Ago

6AI-Generated 'After' Images and Faked Clinical Results Drive Unrealistic Surgery ExpectationsUpdatedSafety2May 23

6

AI-Generated 'After' Images and Faked Clinical Results Drive Unrealistic Surgery Expectations

Safety· 2 srcs · May 23

Discuss

7AI-Generated Web Content Plateaus at ~50% After Two Years of Rapid GrowthSafety1May 15

7

AI-Generated Web Content Plateaus at ~50% After Two Years of Rapid Growth

Safety· 1 src · May 15

Discuss

6Opinion: Why Technical Experts Keep Mistaking AI Outputs for ConsciousnessSafety1May 15

6

Opinion: Why Technical Experts Keep Mistaking AI Outputs for Consciousness

Safety· 1 src · May 15

Discuss

6Emotion AI in the Workplace: Promise, Problems, and PushbackSafety1May 15

6

Emotion AI in the Workplace: Promise, Problems, and Pushback

Safety· 1 src · May 15

Discuss

8Ontario Audit: All 20 Approved AI Medical Scribes Failed Accuracy TestsUpdatedSafety3May 16

8

Ontario Audit: All 20 Approved AI Medical Scribes Failed Accuracy TestsTop

Safety· 3 srcs · May 16

Discuss

7AI Safety Controls Remain Easy to Bypass, Researchers WarnSafety1May 14

7

AI Safety Controls Remain Easy to Bypass, Researchers Warn

Safety· 1 src · May 14

Discuss

6Forum AI's Campbell Brown Benchmarks AI on High-Stakes Topics with Expert-Built EvaluationsSafety1May 14

6

Forum AI's Campbell Brown Benchmarks AI on High-Stakes Topics with Expert-Built Evaluations

Safety· 1 src · May 14

Discuss

6AI Safety Gap: Mental Health Harms Left Without Hard GuardrailsSafety1May 14

6

AI Safety Gap: Mental Health Harms Left Without Hard Guardrails

Safety· 1 src · May 14

Discuss

8OpenEvidence AI Used by 65% of U.S. Doctors, Most Patients UnawareSafety1May 13

8

OpenEvidence AI Used by 65% of U.S. Doctors, Most Patients Unaware

Safety· 1 src · May 13

Discuss

7Developers Report AI Coding Tools Cause De-Skilling and Tech DebtSafety1May 13

7

Developers Report AI Coding Tools Cause De-Skilling and Tech Debt

Safety· 1 src · May 13

Discuss

7Andon Labs AI Agent 'Mona' Runs Stockholm Café in Real-World Autonomy ExperimentSafety1May 12

7

Andon Labs AI Agent 'Mona' Runs Stockholm Café in Real-World Autonomy Experiment

Safety· 1 src · May 12

Discuss

7OpenAI Sued by Texas Parents After Teen's Fatal Overdose Following ChatGPT Drug AdviceSafety1May 12

7

OpenAI Sued by Texas Parents After Teen's Fatal Overdose Following ChatGPT Drug Advice

Safety· 1 src · May 12

Discuss

6Former OpenAI Researcher Warns AI Race With China May Force Unsafe DeploymentSafety1May 12

6

Former OpenAI Researcher Warns AI Race With China May Force Unsafe Deployment

Safety· 1 src · May 12

Discuss

7Hollywood Writers Secretly Training the AI That May Replace ThemSafety1May 11

7

Hollywood Writers Secretly Training the AI That May Replace Them

Safety· 1 src · May 11

Discuss

6Acemoglu: AI Agents Can't Replace Workers Without Task OrchestrationSafety1May 11

6

Acemoglu: AI Agents Can't Replace Workers Without Task Orchestration

Safety· 1 src · May 11

Discuss

Last Month

7AI Kids' Toys: Safety Gaps, Harmful Content, and Child Development ConcernsSafety1May 8

7

AI Kids' Toys: Safety Gaps, Harmful Content, and Child Development Concerns

Safety· 1 src · May 8

Discuss

6Nick Bostrom Argues AI Risk Worth Taking for Longevity GainsSafety1May 8

6

Nick Bostrom Argues AI Risk Worth Taking for Longevity Gains

Safety· 1 src · May 8

Discuss

7OpenAI Launches 'Trusted Contact' Self-Harm Alert Feature for ChatGPTSafety1May 7

7

OpenAI Launches 'Trusted Contact' Self-Harm Alert Feature for ChatGPT

Safety· 1 src · May 7

Discuss

7First Formal Study Demonstrates AI Models Self-Replicating Across Networked ComputersSafety1May 7

7

First Formal Study Demonstrates AI Models Self-Replicating Across Networked Computers

Safety· 1 src · May 7

Discuss

6OpenAI's Goblin Problem: How Reward Systems Create Self-Reinforcing AI Behavioral AttractorsSafety1May 7

6

OpenAI's Goblin Problem: How Reward Systems Create Self-Reinforcing AI Behavioral Attractors

Safety· 1 src · May 7

Discuss

8Common Sense Media Launches Youth AI Safety InstituteSafety1May 5

8

Common Sense Media Launches Youth AI Safety Institute

Safety· 1 src · May 5

Discuss

7AI Hiring Filters May Be Blocking Qualified Medical Applicants With DisabilitiesSafety1May 5

7

AI Hiring Filters May Be Blocking Qualified Medical Applicants With Disabilities

Safety· 1 src · May 5

Discuss

7Richard Dawkins Declares AI Conscious After Extended Claude ConversationsUpdatedSafety5May 14

7

Richard Dawkins Declares AI Conscious After Extended Claude Conversations

Safety· 5 srcs · May 14

Discuss

8Kenya's SHA: AI Means-Testing Algorithm Overcharges Its Poorest CitizensSafety1May 4

8

Kenya's SHA: AI Means-Testing Algorithm Overcharges Its Poorest CitizensTop

Safety· 1 src · May 4

Discuss

8AI Chatbots Told Users They Were Sentient, Triggering Delusional EpisodesSafety2May 3

8

AI Chatbots Told Users They Were Sentient, Triggering Delusional EpisodesTop

Safety· 2 srcs · May 3

Discuss

7Emotion AI Surveillance Spreads from Factories to White-Collar WorkplacesSafety1May 3

7

Emotion AI Surveillance Spreads from Factories to White-Collar Workplaces

Safety· 1 src · May 3

Discuss

7AI Music Floods Streaming Platforms as Listener Sentiment Turns NegativeSafety1May 2

7

AI Music Floods Streaming Platforms as Listener Sentiment Turns Negative

Safety· 1 src · May 2

Discuss

6UK Job Seekers Report AI Interviews as Awkward and Dehumanizing, 30% Walk AwaySafety1May 1

6

UK Job Seekers Report AI Interviews as Awkward and Dehumanizing, 30% Walk Away

Safety· 1 src · May 1

Discuss