Disempowerment potential appeared most often in conversations about relationships & lifestyle or healthcare & wellness—topics where users are most personally invested. Technical domains like software development, which make up ~40% of usage, carried minimal risk.
ETHICS
-
Claude interactions show rare disempowerment risks in vulnerable users
By
–
Over 1.5M Claude interactions, severe disempowerment potential was rare, occurring in 1 in 1,000 to 1 in 10,000 conversations, depending on domain. All four amplifying factors were associated with higher disempowerment rates—but user vulnerability had the strongest effect.
-
Three Ways AI Interactions Can Disempower Users
By
–
We identified three ways AI interactions can be disempowering: distorting beliefs, shifting value judgments, or misaligning a person’s actions with their values. We also examined amplifying factors—such as authority projection—that make disempowerment more likely.
-
Anthropic Research: Disempowerment Patterns in AI Assistant Interactions
By
–
New Anthropic Research: Disempowerment patterns in real-world AI assistant interactions. As AI becomes embedded in daily life, one risk is it can distort rather than inform—shaping beliefs, values, or actions in ways users may later regret. Read more:
-
Clinical Trust in AI: Evidence, Evaluation, and Transparency Requirements
By
–
Forbes gathers perspectives from health leaders on alignment, transparency, and responsible tech to reduce healthcare costs.
— Catherine Adenle (@CatherineAdenle) 28 janvier 2026
It includes a contribution from Elsevier’s Jan Herzhoff on why clinicians' trust in AI requires evidence-based tools, regular evaluation, and transparent… pic.twitter.com/uPgWANJVpiForbes gathers perspectives from health leaders on alignment, transparency, and responsible tech to reduce healthcare costs.
It includes a contribution from Elsevier’s Jan Herzhoff on why clinicians' trust in AI requires evidence-based tools, regular evaluation, and transparent -
Synthetic Data Protects Privacy in AI Development
By
–
On #DataPrivacyDay, we're taking a closer look at synthetic data and how it can keep the AI machine running without sacrificing personally privacy. A tremendous advantage for organizations using synthetic data is that it built, in part, for data privacy — there's little to no
-
CooperBench: AI Agents Perform 50% Worse in Teams
By
–
Introducing the curse of coordination. Agents perform 50% worse in teams than working alone. People building human-AI collaboration today don't realize why current LLMs fail to be good teammates. We built CooperBench to study this. For humans, we recognize that teamwork isn't just the sum of individual capability. Communication and coordination often outweigh raw skill. But for AI? We're only hill-climbing benchmarks that evaluate solo technical abilities. CooperBench A benchmark to evaluate agent cooperation in realistic software teamwork tasks. The setup is intuitive: two agents, two tasks, two VMs, one chat channel (agents can send over arbitrary text, even the entire patch they wrote). We evaluate whether the merged solution from both agents passes the requirements of both tasks. The curse of coordination The most striking result: agents perform 50% worse in teams (black line) than working alone (blue line). Why is this happening? Is it because they can't use the communication tool? No. They spent 20% of their time sending messages. The problem? Those messages were repetitive, vague, ignored questions, or straight-up hallucinated. But bad communication is only part of the story. We found two deeper failures: Commitment: Agents don't do what they promised. Expectations: Agents don't expect others to keep promises either. Without these, cooperation collapses. However, there is a silver lining We also find emergent coordination behaviors, e.g. role division, resource division, and negotiation, which gives us hope that we can use reinforcement learning to improve coordination. What's next? It is true that highly-engineered multi-agent orchestration could largely sidestep the coordination problem. However, we care more about the AI's capability: if we truly want AI to be our teammates, we need them to be natively capable of effective communicating and coordinating. Two agents on software tasks is just the beginning. The real goal: agents that can cooperate with us well enough to actually empower us. CooperBench is our first step. If you're working on this too, let's talk.
-
AI Leaders Panel on Transparency and Accountability in Advanced Analytics
By
–
Final Call! Join a hand-picked panel of AI and Risk leaders for a deep dive into transparency, accountability, and value delivery in advanced analytics. Today at 10 AM PT | 1 PM ET Last chance to register: https://
domino.buzz/4t6tyO2 -
AI Adoption Scaling Faster Than Organizational Foundations
By
–
Trust in AI is rising, and that sounds like progress, until you look at how many organizations are scaling faster than their foundations. Informatica’s CDO Insights 2026 report captures a reality I am seeing everywhere. AI adoption is moving from pilots to everyday operations,
-
AI in question: everything is pre-recorded and parameterized
By
–
L’intelligence artificielle en question.
— Defend Intelligence (Anis Ayari) (@DFintelligence) 28 janvier 2026
Non en vrai ça fait de la peine à ce niveau. Ils enregistrent un vieux fichier mp4 qui bouge. Il le le projette et le font passer pour une IA critique qui pose des questions qui « fâchent » alors que tout est pré enregistré et paramétrer.… https://t.co/aacYCUf1ma pic.twitter.com/tTHEzcbS6IThe artificial intelligence in question. No, really, it's pathetic at this level. They record an old mp4 file that moves. They project it and pass it off as a critical AI that asks 'tough' questions, while everything is pre-recorded and parameterized.