Trying to build some authentication protocols into it in collaboration with a bunch of Stanford folks. Very fun space!
13.02.2025 19:15 β π 1 π 0 π¬ 1 π 0@tobinsouth.bsky.social
PhD student @MIT & @medialab; Building a secure & robust computational future; π¦πΊπ¦; Fulbrighter; he/him
Trying to build some authentication protocols into it in collaboration with a bunch of Stanford folks. Very fun space!
13.02.2025 19:15 β π 1 π 0 π¬ 1 π 0π Read the full paper (examples, motivations, legal insights, technical designs, and more): arxiv.org/pdf/2501.09674
27.01.2025 17:55 β π 1 π 0 π¬ 0 π 0Amazing to collaborate with Samuele Marro, Thomas Hardjono, Robert Mahari, @cedricwhitney.bsky.social, Dazza Greenwood, Alan Chan, Alex Pentland, and many others providing feedback.
27.01.2025 17:55 β π 0 π 0 π¬ 1 π 0Weβve even tackled natural language permissions!
Our framework translates human-readable instructions into auditable, structured access controls.
Think: "Only read project files, avoid financial folders" β enforceable, machine-readable rules. π
Delegation isn't just technical. It's crucial for:
β
Protecting human-only spaces online π
β
Preventing misuse in high-stakes domains (finance, healthcare) π₯π³
β
Enabling safe, auditable agent collaboration π οΈ
Our framework builds on existing web standards (OAuth 2.0, OpenID Connect) to:
β
Identify agents uniquely
β
Limit permissions & scope
β
Maintain audit trails
β
Ensure human oversight
All while maintaining compatibility with web infrastructure. π
Autonomous AI agents are growing fastβhow do we ensure they act responsibly, securely, and transparently?
OpenAI Operator isn't good enough...
π Authenticated Delegation is our answer:
A secure way to delegate authority while maintaining accountability and control.
How can we build better delegation to AI agents?
π¨ New Paper on *Autheticated Delegation* for AI Agents. Letβs make human-in-the-loop systems work for people.
Thank you @melissahei.bsky.social for covering the Data Provenance Initiative's new study on π Multimodal data π.
Our study finds:
- A concentration of data power
- Western-centricity isn't improving for 10+ years
Read more! www.dataprovenance.org/Multimodal_D...
At this point, I'm just racing to write my PhD thesis before o4 comes out and would do a better job
20.12.2024 19:25 β π 2 π 0 π¬ 0 π 0The cursor team really can't catch a break
20.12.2024 19:25 β π 1 π 0 π¬ 0 π 0@klyman.bsky.social is hands down the funnest Evals talk yet.
15.12.2024 22:55 β π 1 π 0 π¬ 0 π 0This is the best abstract I've perhaps ever seen...
02.12.2024 23:01 β π 0 π 0 π¬ 0 π 0I always think not eating until the feast is a smart move until I arrive and someone starts handing me French 75s
29.11.2024 18:31 β π 1 π 0 π¬ 0 π 0@mbakker.bsky.social is the Troye Sivan of AI
26.11.2024 21:13 β π 1 π 0 π¬ 0 π 0I'm glad I have access to a new app that I can lurk in and post once a year
24.11.2024 19:02 β π 4 π 0 π¬ 1 π 0