E-commerce fraudulent websites are a big problem, yet the community lacks open and comprehensive approaches to proactively discover them. In our upcoming NDSS paper we present Loki, a system that mines search engine queries to surface new scam websites. Read our paper here: arxiv.org/pdf/2509.12181

Spot the differences

Practicing for my plan B career

Worthy companion

Also this paragraph might apply here sadly

In the wake of recent funding cuts, maybe I should change career. Thanks targeted ads!

That's odd, it refuses to create images for the same prompt for me. I noticed that it's very temperamental, for sure the safety mechanisms aren't bulletproof...

Spending the week in the beautiful Brazilian countryside brainstorming research collaborations

The NSF Staff Directory page seems to be gone...

Genius piece of work.

ACSAC2024 was a success! Thanks to everyone who made the conference possible and see you next year! In case you missed this edition, we’ll still be in Waikiki!

Almost ready to kick off the @acsacconf.bsky.social main program tomorrow!

This resonates with our research where we found that Russian trolls on Reddit had entire conversations among each other, taking opposite sides with the goal of polarizing discussion, with the goal of shifting the vibe and conversation topics of entire subreddits.

📝 arxiv.org/pdf/2112.00443

Post a pic YOU took (no description) to bring some zen to the timeline

Check out our USENIX Security paper "PIXELMOD: Improving Soft Moderation of Visual Misleading Information on Twitter". We combine perceptual hashing, vector databases, and OCR in a scalable and efficient pipeline to flag image-based misinformation on social media.

seclab.bu.edu/people/gianl...

Later this month my student Cici Ling will present our work on studying image-based COVID misinformation on Twitter at CSCW 2023. She will be on the faculty job market this year, make sure to talk to her if you are hiring!

You can read our paper here: seclab.bu.edu/people/gianl...