I notice that a journalist picked up the info and that many of you are keen to let me know how much you don't give a shit -- okay! I took screenshots of the most absurd blocklists and not all of them, but I'm not here to explain who I am for the umpteenth time because, again, who cares?

Giada Pistilli@giadapistilli.com · 9mo ago

I really wanted this social network to work for me. But apparently, I’m on the worst block lists out there… a bit unfair. If you are interested in my research, follow me on LinkedIn.

I really wanted this social network to work for me. But apparently, I’m on the worst block lists out there… a bit unfair. If you are interested in my research, follow me on LinkedIn.

Il mio nuovo op-ed su @wired.com parte da un dato inquietante: milioni di persone oggi confidano le proprie crisi emotive all’intelligenza artificiale. Non perché la credano umana. Ma perché non trovano alternative umane disponibili.

ChatGPT e la società del conforto artificiale. L'AI sta diventando il nuovo specchio della solitudine

Dopo l'ammissione da parte di OpenAI che una parte di utenti confida al suo chatbot segni di gravi disturbi e attaccamenti emotivi, emerge ancora più chiaramente il problema dell'attaccamento a questi...

wired.it

Join the @ellisinstitute.fi Scientific Seminar in Espoo 🇫🇮 on Nov 18: a landmark event uniting research, industry and policy. Speakers: @giadapistilli.com, Kyunghyun Cho, @ericmalmi.bsky.social, @serge.belongie.com, Max Welling, @lrlaurahy.bsky.social 🔗 www.ellisinstitute.fi/launch-seminar

Map of the internet: 1.3M nodes (BGP)

What if your most personal chat logs became the next source of ad data? @frimelle.bsky.social and I wrote an op-ed for @techpolicypress.bsky.social We look at what happens when generative AI conversations (the ones we treat as private) are turned into raw material for targeted advertising.

Before AI Exploits Our Chats, Let’s Learn from Social Media Mistakes | TechPolicy.Press

Privacy in the age of conversational AI is a governance choice, write Hugging Face's Lucie-Aimé Kaffee and Giada Pistilli.

techpolicy.press

🌎 AI ethics and sustainability are two sides of the same coin. In our new blog post with @sashamtl.bsky.social, we argue that separating them (as is too often the case) means missing the bigger picture of how AI systems impact both people and the planet.

Ethics + Sustainability = Responsible AI

A Blog post by Sasha Luccioni on Hugging Face

huggingface.co

À tout à l’heure !

La Science, CQFD@sciencecqfd.bsky.social · 10mo ago

Et demain, on consacrera notre émission aux conséquences potentiellement délétères des IA sur la santé mentale 🤖 Ce sera à 16h sur @franceculture.fr ! #ScienceCQFD

One of the hardest challenges in AI safety is finding the right balance: how do we protect people from harm without undermining their agency? This tension is especially visible in conversational systems, where safeguards can sometimes feel more paternalistic than supportive.

I've noticed something. While we're careful about what we post on social media, we're sharing our deepest and most intimate thoughts with AI chatbots -- health concerns, financial worries, relationship issues, business ideas... huggingface.co/blog/giadap/...

Advertisement, Privacy, and Intimacy: Lessons from Social Media for Conversational AI

A Blog post by Giada Pistilli on Hugging Face

huggingface.co

With @frimelle.bsky.social and @yjernite.bsky.social, we released the AI Companionship Leaderboard on @hf.co to see how models handle connection, intimacy, and vulnerability. huggingface.co/spaces/frime...

Companionship Leaderboard - a Hugging Face Space by frimelle

Browse and analyze benchmark data for different language models. View metrics like Average, Assistant Traits, and Emotional Investment. Customize the columns to display and search for specific models.

huggingface.co

L’IA va-t-elle bouleverser l’enseignement et la place des professeurs ? Hier à 18h20 sur @franceculture.fr, @quentinlafay.bsky.social recevait @ccailleaux.bsky.social‬, professeur d’histoire, @giadapistilli.com, spécialiste de l’éthique en IA et Orianne Ledroit, DG d’EdTech France

L’IA va-t-elle bouleverser l’enseignement et la place des professeurs ?

À quelques jours de la rentrée scolaire, Elisabeth Borne a annoncé mardi 26 août 2025, la mise en place d’une intelligence artificielle pour « accompagner les professeurs dans leur métier ».

radiofrance.fr

🚨 Releasing INTIMA (Interactions and Machine Attachment Benchmark): an evaluation framework for measuring how AI systems handle companionship-seeking behaviors. huggingface.co/papers/2508.... Thread on what we discovered, together with @frimelle.bsky.social and @yjernite.bsky.social

Paper page - INTIMA: A Benchmark for Human-AI Companionship Behavior

Join the discussion on this paper page

huggingface.co

🚨 Releasing INTIMA (Interactions and Machine Attachment Benchmark): an evaluation framework for measuring how AI systems handle companionship-seeking behaviors. huggingface.co/papers/2508.... Thread on what we discovered, together with @frimelle.bsky.social and @yjernite.bsky.social

Paper page - INTIMA: A Benchmark for Human-AI Companionship Behavior

Join the discussion on this paper page

huggingface.co

From Replika to everyday chatbots, people form emotional bonds with AI. But what happens when an AI tells you "I understand how you feel" and you actually believe it? With @frimelle.bsky.social and @yjernite.bsky.social, we dug into something: how AI systems handle our emotional lives.

AI Companionship: Why We Need to Evaluate How AI Systems Handle Emotional Bonds

A Blog post by Giada Pistilli on Hugging Face

huggingface.co

I am so grateful that this paper is out! It so eloquently describes so many of the issues I've seen with a lot of AGI experimentation that concludes that LLMs are scheming, ill-intentioned, or malicious, and the chimp language metaphor is the cherry on the cake 🍰 arxiv.org/abs/2507.03409

Lessons from a Chimp: AI "Scheming" and the Quest for Ape Language

We examine recent research that asks whether current AI systems may be developing a capacity for "scheming" (covertly and strategically pursuing misaligned goals). We compare current research practice...

arxiv.org

I've been posting bits about this research, but now: new paper alert! 🚨 My colleague @brunatrevelin.bsky.social and I just shared research on why traditional consent frameworks are breaking down in AI contexts.

Can AI Be Consentful? Rethinking Permission in the Age of Synthetic Everything

A Blog post by Giada Pistilli on Hugging Face

huggingface.co

AI is making everyone on dating apps sound charming. But is it still you talking? When we start outsourcing our most intimate expressions, we risk losing authenticity. I spoke to @washingtonpost.com about the rise of chatbots in dating and why it may feel good, but erodes emotional honesty.

‘It’s almost like we never even spoke’: AI is making everyone on dating apps sound charming

Dating apps say AI can help people who are too busy, shy or abrasive to win dates. But a growing number of singles are finding that AI makes dating more complicated.

washingtonpost.com

Avec @natachatriou.bsky.social - est-elle cyborg?- et la @sciencecqfd.bsky.social , faites un vrai tour d'horizon de l'#IA : à quand de l'IA faible à l'IA générale, qui pourrait surpasser l'humain ? avec @quecalcoatle.bsky.social @giadapistilli.com #AxelCypel de vrais humanoïdes !

De l'IA faible à l'IA forte : la paranoïa : épisode 3/5 du podcast IA, technosciences, demain leur appartient

AUDIO • IA, technosciences, demain leur appartient, épisode 3/5 : De l'IA faible à l'IA forte : la paranoïa. Une série inédite proposée par France Culture. Écoutez La Science, CQFD, et découvrez nos p...

radiofrance.fr

[RDV 16h !] L’IA générale, qui pourrait surpasser l'humain, fait couler beaucoup d’encre. Il paraît même que c’est pour cette année : toutefois, a-t-on une définition commune de ce qu’elle pourrait être ? tinyurl.com/28b32lpm avec @quecalcoatle.bsky.social, @giadapistilli.com et Axel Cypel

Hi! I’ll be in London July 9–11. Who’s around to chat about current AI research?

LM/ML models are being developed to detect neurodev & psychiatric conditions from speech. Are we building the right things? What are we overlooking? We built a framework rooted in ethics & real-world needs: ojs.aaai.org/index.php/AI... w @rockberta.bsky.social @giadapistilli.com Maheswari 1/

🗣️ Whose voice do we hear when AI speaks? Every language carries its own cultural values and worldviews. So, when we build AI systems, we're not just deciding how they speak but also whose perspectives they represent. huggingface.co/blog/giadap/...

Whose Voice Do We Hear When AI Speaks?

A Blog post by Giada Pistilli on Hugging Face

huggingface.co

The sycophancy problem in one line, courtesy of @giadapistilli.com Really interesting piece by @melissahei.bsky.social www.ft.com/content/72aa...

Have some of you researched why LLMs tend to give themselves "astral names"? Like Nova, Luna, Sol... I've seen this across different closed and open-weight models and scenarios.

Ever notice how some AI assistants feel like tools while others feel like companions? Turns out, it's not always about fancy tech upgrades, because sometimes it's just clever design. huggingface.co/blog/giadap/...

AI Personas: The Impact of Design Choices

A Blog post by Giada Pistilli on Hugging Face

huggingface.co

Ever notice how some AI assistants feel like tools while others feel like companions? Turns out, it's not always about fancy tech upgrades, because sometimes it's just clever design. huggingface.co/blog/giadap/...

AI Personas: The Impact of Design Choices

A Blog post by Giada Pistilli on Hugging Face

huggingface.co

Very proud of this work led by my amazing colleague @mmitchell.bsky.social! Read her interview for Wired here: www.wired.com/story/ai-bia...

AI Is Spreading Old Stereotypes to New Languages and Cultures

Margaret Mitchell, an AI ethics researcher at Hugging Face, tells WIRED about a new dataset designed to test AI models for bias in multiple languages.

wired.com

Margaret Mitchell@mmitchell.bsky.social · last yr.

Many thanks to @wired.com & @thiccreese.bsky.social for covering our work on stereotype bias!! The new dataset is SHADES, now available here: huggingface.co/datasets/LanguageShades/BiasShades The first of its kind for analyzing GenAI stereotyping across many languages+cultures. Published at NAACL.🤗

I spoke to @washingtonpost.com about how bad actors are "grooming" LLMs with disinformation. Most chatbots struggle with disinformation, which worsens with search-augmented systems prioritizing recent information. The more real-time these models become, the more vulnerable they are to manipulation.

Russia seeds chatbots with lies. Any bad actor could game AI the same way.

In their race to push out new versions with more capability, AI companies leave users vulnerable to “LLM grooming” efforts that promote bogus information.

washingtonpost.com

🤗 New from us! Just published a blog post exploring how we're rethinking consent in the AI ecosystem. Here's what we're seeing in the @hf.co Hub that differs from traditional closed systems...

Consent by Design: Approaches to User Data in Open AI Ecosystems

A Blog post by Giada Pistilli on Hugging Face

huggingface.co