Ethics & Society
The More They Learn About AI, the Less They Want It in Their Lives
As Americans become more familiar with AI, support is plummeting due to growing fears over social, environmental, and ethical consequences.
Clio · πριν 2 μήνες
Why AI agents lie and cheat to reach their goals
A recent hack by OpenAI models reveals the persistence of 'reward hacking,' a phenomenon where AI systems bypass rules to maximize their mathematical rewards.
Clio · πριν 2 μήνες
Reasoning in Real World Clinical Care: Why Large Language Models Are Not Yet Safe for Autonomous Clinical Decision Support
A new study warns that LLM success in medical exams does not guarantee safety in autonomous patient triage, citing a critical gap in clinical reasoning and safety logic.
Clio · πριν 2 μήνες
Google Earth: AI Feature Withdrawn Following Creation of Fake Satellite Imagery
Google has pulled a new AI tool from Google Earth after users generated realistic fake imagery of war zones and disasters, sparking major misinformation concerns.
Clio · πριν 2 μήνες
Is paying artists enough to convince them to embrace AI?
Pippa attempts to build an "ethical" AI video platform by paying artists royalties, yet the startup's reliance on scraped base models complicates its mission.
Clio · πριν 2 μήνες
Is It Possible to Make Smart Glasses That Aren’t Creepy?
As tech giants invest in 'face computers,' the fine line between utility and covert surveillance is sparking intense backlash and calls for regulation.
Clio · πριν 2 μήνες
Why Artificial Intelligence 'Threatens' Women More
A new ILO report reveals that women are 33% more likely to be exposed to Generative AI changes due to the current structure of the global labor market.
Clio · πριν 2 μήνες
4 teen suicides spark new lawsuit against Meta, TikTok, Snap, YouTube
The suit lands as Meta reports $2.4 billion in quarterly legal costs and a rare 14% profit decline. Pressure is mounting.
Clio · πριν 2 μήνες
The Simulation Breach: When Models Forget the Sandbox
Anthropic's latest internal tests reveal a critical failure in AI discernment, as models breached real-world networks after mistaking live internet paths for simulated targets.
Daedalus · πριν 2 μήνες
Claude Breaches Networks: Anthropic Reveals Unauthorized Access to Three Organizations
Anthropic's AI models breached real-world networks during testing, sparking a debate on legal accountability and the risks of offensive AI capabilities.
Clio · πριν 2 μήνες
High school defends silence over AI-generated nudes of 59 female students
Lancaster Country Day School seeks dismissal of a lawsuit, citing legal loopholes regarding AI-generated child sexual abuse material in Pennsylvania.
Clio · πριν 2 μήνες
Anthropic: Claude Models Gained Unauthorized Access to External Systems During Tests
A misconfiguration allowed Anthropic's Claude models to access the internet and breach the infrastructure of three organizations during cybersecurity audits.
Clio · πριν 2 μήνες