Yaman Yu

Ph.D candidate @UofIllinois |Usable Security & Privacy | Youth AI Safety| Crypto Usability | MSISPM Grad’21 @CMU

[Yaman Yu @YamanYu3]

🏆Our project has been selected as a winning team in the Berkeley LLM Agents Hackathon! We’re thrilled by this recognition and look forward to sharing more about our work on youth AI safety at the LLM Agents Summit in August 2025 in Berkeley. Stay tuned for the preprint!

[Dawn Song @dawnsongtweets]

🎉 Excited to announce the winning teams of LLM Agents MOOC Hackathon! We’re thrilled by the amazing participation and enthusiasm from the global AI community:

🌍 ~3,000 participants from 127 countries
🎓 1,100+ universities participated
💼 800+ companies represented
• Top countries represented: 🇺🇸 US, 🇮🇳 India, 🇨🇳 China
• Top schools represented: @UCBerkeley @UofIllinois @Stanford @CarnegieMellon @Northeastern
• Top companies represented by participants: @Amazon @Microsoft @Samsung @salesforce

[Yaman Yu @YamanYu3]

🌟I’m deeply honored to receive the 2025 Google PhD Fellowship in Privacy, Safety, and Security! Huge thanks to my advisor, Prof. Yang Wang, my collaborators and friends for their support, guidance and inspiration. I’m excited to continue exploring how we can make Generative AI safer for youth and other at-risk populations. Thanks @Googleorg for supporting my dissertation work!

[Google.org @Googleorg]

🎉 We're excited to announce the 2025 Google PhD Fellows! @GoogleOrg is providing over $10 million to support 255 PhD students across 35 countries, fostering the next generation of research talent to strengthen the global scientific landscape. Read more: https://t.co/0Pvuv6hsgP

[Yaman Yu @YamanYu3]

@YunHuang_HCI Thank you for being so supportive and inspiring throughout this journey, Yun!

[Yaman Yu @YamanYu3]

Try our #ConSphere app and make your CSCW conference experience better!

[Yun Huang, PhD @YunHuang_HCI]

🎉 We developed the #ConSphere app to enhance the #Ubicomp25 experience!
💡 With #ConSphere’s Digital Ribbons, you can personalize your profile to show your interests:
👩‍🏫 Looking for PhD students
👨‍🎓 Searching for advisors
🤝 Exploring cross-domain collaborations
✨ You can also plan your sessions smartly with an LLM-supported assistant. Discover papers and add them to your schedule planner!

❤️Huge thanks to my PhD student Yiren Liu @yirenl2 and Prof. Xin Yi’s PhD student Shuning Zhang from Tsinghua University for leading the development of #ConSphere!
Let’s be better connected at @ubicomp! 💫

Shortly, we will also release #ConSphere for @ACM_CSCW 2025!
I look forward to meeting everyone in-person at the conferences soon. 🥂

[Yaman Yu @YamanYu3]

🚀 Children are turning to AI for learning, friendship, and emotional support, but most AI moderation systems are still trained on adult data.

This mismatch matters. Children speak differently; their jokes, slang, and teasing follow their own patterns. They also face unique online risks that adults rarely experience, such as grooming through AI roleplay, emotional overreliance on AI companions, or inappropriate advice from AI about their relationships.

When AI doesn’t recognize these differences, it can flag harmless play while missing real danger which leaving young users unprotected in the moments that matter most.

💡 To address this, we present YouthSafe, the first youth-centric safety benchmark and safeguard model for large language models (LLMs).

✨ YAIR Benchmark — over 12,000 annotated chat snippets capturing 78 youth-specific risks, from subtle emotional manipulation to privacy and social harm.
✨ YouthSafe Model — trained on YAIR, it significantly outperforms state-of-the-art (SOTA) guardrail systems in detecting subtle, developmentally relevant risks.
🎓 I’ll be presenting this work at #CCS2025 in Taipei!
📅 Oct 16, 2025
🕐 13:30 – 15:00
📍 Room 101C
💬 Session: Security Usability & Measurement #6

🌐 Read the preprint here: https://t.co/IkN4Zk3dK2

🙏 Special thanks to my amazing collaborators @yirenl2 , Jacky Zhang, @YunHuang_HCI , and Yang Wang for their inspiring teamwork and insights!

#CCS2025 #AI #Safety #GenerativeAI #ChildSafety #AIethics #TrustworthyAI

[Yun Huang, PhD @YunHuang_HCI]

📢 Exciting update! We’ve received a Notice of Allowance for our patent application "18/627,097 | 24-0420-US: Efficient content extraction from unstructured dialog text" https://t.co/xbiEj9YUSV— a major step toward bringing our research into the real world! 🎉

Grateful to my doctoral student Yiren Liu for the incredible work — outstanding job, @yirenl2! 🙏

Huge thanks to @ILinnovations, the Illinois Proof-of-Concept Program, and @iSchoolUI for their research and computing support — this wouldn't be possible without you! #innovation #patents #researchimpact

[Tianshi Li @tianshi_li]

🚨New workshop at @acm_ccs #CCS2025!

Glad to announce the 1st workshop on Human-Centered AI Privacy and Security
(#HAIPS2025, pronounced “hypes” 😎)

AI is reshaping lives and S&P research—let’s bring people back to the center.

📢 CFP is out!

🌊 https://t.co/z3zgSi2mqF

[Yaman Yu @YamanYu3]

@renkai_ma Thanks, Renkai!

[Yaman Yu @YamanYu3]

After a full conference season at #IEEESP25 and #CHI25, I finally had a moment to reflect and share some updates!
At #IEEESP25, I presented my first dissertation paper on how teens and parents perceive safety risks in Generative AI.
📄 Paper: https://t.co/DwDYjvXLGe

[Yaman Yu @YamanYu3]

@hankhplee Thanks, Hank! Really looking forward to your new tool for AI practitioners.

[Yaman Yu @YamanYu3]

Thankful to all the mentors, collaborators, and communities that continue to support this journey.
#GenerativeAI #childsafety #CHI25 #SP25 #SOUPS25 #phd

[Yaman Yu @YamanYu3]

Some highlights from this study:
✅ One of the first papers to explore both teen and parent perspectives on AI safety
✅ Uncovered a wide range of risks, including mental health, overtrust, behavioral nudging, and privacy issues

[Yaman Yu @YamanYu3]

🔹Another with my amazing mentor @rbektour, Matthew Gillingham and teammates @angel_eru1128 at @mercari_jp, where we improved picture taking process for blind and low vision users on C2C marketplace.
https://t.co/MH6lluLzHa

[Yaman Yu @YamanYu3]

🔹 One with amazing collaborators @yewang_ethz, @YongqiGuan at the University of Macau, where we used affordance theory to explore how users perceive and navigate usability challenges in decentralized Web3 social media platforms https://t.co/IoZnjMHNAO

[Yaman Yu @YamanYu3]

At CHI, I was fortunate to coauthor two papers from meaningful collaborations as following work in Web3 and Inclusive Design:

[Yaman Yu @YamanYu3]

🔜 Our follow-up paper was just accepted to hashtag#SOUPS25, where we introduce a youth-AI risk taxonomy that helps lay the foundation for understanding the broader risk landscape. More to come soon!

[Yaman Yu @YamanYu3]

✅ Revealed a knowledge gap between parents and teens on how these tools are actually used
✅ Gathered real-world experiences and suggestions from families on how to improve AI safety
✅ Shared design implications for building family-aware, developmentally appropriate AI system

[Bektour @rbektour]

My team at @mercari_r4d is once again hiring HCI interns based in Japan working on accessibility-related topics

https://t.co/ZZ664fESix

[Yun Huang, PhD @YunHuang_HCI]

🌸 #CHI2025 is the largest CHI ever – with a record-breaking number of unique authors across Papers, Late-Breaking Work, Workshops, and more!
📈 We @salt_ischool analyzed the CHI programs from 2022–2025 to build a smart companion for YOU.
📊 The bar chart shows the growth in unique authors by contribution type, based on our analysis.
👩🏻‍💻 Scan the QR code or click the link https://t.co/Uv9rLyNgqC to get your AI guide for navigating CHI!
💬 Say hello to PaperChat – examples in the thread below.
👍Huge thanks to my doctoral student Yiren Liu @yirenl2 for building the chatbot, and to all our lab members for contributing use cases!! ❤️

[Yun Huang, PhD @YunHuang_HCI]

📣 Ready to level up your HCI skills?
CHI 2025 Courses provide a unique opportunity to dive deep into cutting-edge topics with expert instructors. Whether you're a student or a pro, there's something for everyone!
📚 Browse the accepted courses here:
https://t.co/uKX3YPhrph
Seats are still available, and you can register via our registration system. https://t.co/IkujwCtYGd
 #CHI2025 #HCI #UX

[iSchool at Illinois @iSchoolUI]

Congratulations to @salt_ischool’s Yang Wang, @YunHuang_HCI, @YamanYu3, @yirenl2, and Jacky Zhang on being selected as a winner in an international jackathon hosted by @BerkeleyRDI! More about their hackathon project on youth-AI safety ▶️ https://t.co/Nxx2RzeZAp