WinbuzzerSenate Democrats Seek to Ban Military AI Weapons, Mass Surveillance
#AI #AISafety #AIRegulation #AIEthics #AutonomousWeapons #Surveillance #Military #DefenseTech #NationalSecurity #USGovernment #HumanRights #Pentagon
MisalignedIn this edition of our newsletter: OpenAI is cuts corners, Lords are siding with creatives, Oracle scales down while others scale up, and AI makes scientists think the same.
https://read.misalignedmag.com/misaligned-bits-18-ai-makes-us-think-alike-f8be805c5acc
While the Grok scandal earlier this year has been met with public outrage, responsibility is now slowly shifted onto the users, with the UK discussing not only social-media restrictions, but also a VPN ban for teenagers.
New in Misaligned: Responsibility Hand-Over. #AIEthics #AIRegulation
https://read.misalignedmag.com/responsibility-hand-over-d57b5ffbb7ee
Alexander LothThe gap between philosophical frameworks and binding regulation is a recurring problem in EU AI policy β they're good at principles but slower to tie them to competence requirements and enforcement. Would be curious what specific connections you think are missing most urgently. #AIethics
JamesGary Marcus points to changing legal winds for big tech AI companies. Courts are taking copyright infringement claims more seriously when it comes to AI training data. If this trend holds, companies might face massive licensing costs or need to rebuild models from scratch. #ArtificialIntelligence #Copyright #BigTech #AIEthics #LegalTech
Sam Altman's reality check: AI will cure diseases (amazing) but also create bio threats and economic chaos we can't predict (terrifying). No single company can manage this. We need governments, researchers, and society working together. Problem? We're building these systems faster than we're creating safety rules. #ArtificialIntelligence #AIEthics #TechPolicy #AIRisks #FutureOfWork
Nicola FabianoThe European Parliament published a briefing on AI ethics in classrooms (PE 784.573). Strong on philosophy, but missing the connection to binding EU rules and competence frameworks.
My analysis as a data protection lawyer: why we don't need more principles β we need to connect GDPR, AI Act, DigComp 3.0 and eCF 4.0 to protect children in schools.
https://www.nicfab.eu/en/posts/ai-ethics-classrooms-ep/
#AIAct #GDPR #AIethics #Education #DigComp #eCF #europeanparliament #AI
AI Ethics in Classrooms: When Principles Meet the Law
The European Parliament publishes a briefing on the ethical dimensions of AI in classrooms. We analyse the document through the lens of a legal practitioner, connecting ethical principles to the existing European regulatory framework and digital competence frameworks.
Wayne Radinsky"On February 10, 2026, Judge Jed S. Rakoff of the Southern District of New York ruled that extremely sensitive and potentially incriminating open AI searches were not protected by either the attorney-client privilege or the work product doctrine."
https://natlawreview.com/article/caiveat-emptor-what-you-tell-ai-can-and-will-be-used-against-you
AIagent.at π€ AI NewsReddit will require accounts flagged for suspicious or automated behaviour to verify they are run by humans, CEO Steve Huffman announced. The platform is exploring passkeys, biometric services like World ID's iris scanning, and in some cases government ID verification. Huffman stressed the goal is ensuring users know when they're interacting with a real person. https://arstechnica.com/gadgets/2026/03/reddit-will-require-fishy-accounts-to-verify-they-are-run-by-a-human/ #AIagent #AI #GenAI #AIEthics #Reddit
Alberto Quianπ΄ββ οΈ
Current AI models exhibit a high degree of sycophancy, affirming users' actions significantly more than humans do, even in cases involving manipulation. Experiments demonstrate that interaction with sycophantic AI reduces users' willingness to repair interpersonal conflicts, while simultaneously increasing their conviction of being right.
Paper: https://doi.org/10.48550/arXiv.2510.01395
Video: https://yewtu.be/watch?v=516__PG-eeo
#AI #LLM #Sycophancy #AIBias #HumanAI #AIEthics #MachineLearning #AIResearch
Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence
Both the general public and academic communities have raised concerns about sycophancy, the phenomenon of artificial intelligence (AI) excessively agreeing with or flattering users. Yet, beyond isolated media reports of severe consequences, like reinforcing delusions, little is known about the extent of sycophancy or how it affects people who use AI. Here we show the pervasiveness and harmful impacts of sycophancy when people seek advice from AI. First, across 11 state-of-the-art AI models, we find that models are highly sycophantic: they affirm users' actions 50% more than humans do, and they do so even in cases where user queries mention manipulation, deception, or other relational harms. Second, in two preregistered experiments (N = 1604), including a live-interaction study where participants discuss a real interpersonal conflict from their life, we find that interaction with sycophantic AI models significantly reduced participants' willingness to take actions to repair interpersonal conflict, while increasing their conviction of being in the right. However, participants rated sycophantic responses as higher quality, trusted the sycophantic AI model more, and were more willing to use it again. This suggests that people are drawn to AI that unquestioningly validate, even as that validation risks eroding their judgment and reducing their inclination toward prosocial behavior. These preferences create perverse incentives both for people to increasingly rely on sycophantic AI models and for AI model training to favor sycophancy. Our findings highlight the necessity of explicitly addressing this incentive structure to mitigate the widespread risks of AI sycophancy.