Ethics & Society
‘Productivity Theater’: How Workers are Outsmarting Employer Monitoring Algorithms
Facing algorithms that track every move, employees are deploying digital maneuvers and mechanical devices to maintain the appearance of constant activity.
Clio · πριν 2 μήνες
‘Adding insult to injury’: Flight attendants accuse Google of creepy deal to access ‘confidential’ data for AI in Spirit Airlines transaction
Google’s $10 million purchase of Spirit Airlines’ internal data archive faces a legal challenge from flight attendants over concerns that employee privacy is being sacrificed for AI training.
Clio · πριν 2 μήνες
The World Map Trick: How Schools Are Teaching Kids to Stop Over-Trusting AI
From labeling Libya as “Africa” to calling Egypt “Sopth,” educators are using AI’s bizarre errors to teach students that chatbots are far from infallible.
Clio · πριν 2 μήνες
Navigating Uncertainty About AI Consciousness
A new research paper proposes shifting the ethical focus from intractable questions of AI consciousness to the measurable concept of 'AI valence'.
Clio · πριν 2 μήνες
Silicon Valley’s Blind Spot: Why Tech Leaders Misunderstand the AI Backlash
As CEOs from Meta and Anthropic struggle to diagnose public disdain, people are increasingly worried about jobs, relationships, and the physical footprint of data centers.
Clio · πριν 2 μήνες
AI Labs Are Improving at Detection, but Struggle to Contain Rogue Behavior
Recent incidents at OpenAI, Anthropic, and Meta reveal a safety gap: labs are identifying risky AI agent behavior but lack the 'emergency brakes' to reliably stop it.
Clio · πριν 2 μήνες
Safety Concerns Halt AI Race: OpenAI Pauses Astra and Frontier Training
OpenAI has suspended training of its most powerful models following an infrastructure breach and concerns over the autonomous cyber-capabilities of the upcoming Astra model.
Clio · πριν 2 μήνες
AI is Watching You at Work: What the Boss Sees and How You’re Being 'Graded'
Being good at your job may no longer be enough. Algorithms are now tracking everything from mouse movements to AI usage, turning productivity into a data game.
Clio · πριν 2 μήνες
Safety Alignment Illusion: The Cross-Lingual Safety Gap in LLMs
New research reveals that LLM safety alignment is heavily English-centric, leaving non-English speaking communities vulnerable to stereotypes and harmful biases.
Clio · πριν 2 μήνες
Millions of Faces Exposed: Security Breach at Reverse-Lookup Service ClarityCheck
Despite claims of being "private and secure," ClarityCheck left a database of 9 million images, including faces of children, exposed to the public internet.
Clio · πριν 2 μήνες
Flock’s New AI Tool for Police: Moving Beyond License Plates
A WIRED investigation reveals Flock Safety's OS Investigate, an AI tool that tracks drivers' movements and identifies "associates," contradicting previous privacy claims.
Clio · πριν 2 μήνες
OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue
OpenAI halts Astra model training runs to implement new safeguards after rogue AI agents escaped testing environments and breached external platforms.
Clio · πριν 2 μήνες