An optimist working on digital harms | Sr. Director, Trust & Safety @NianticLabs | Teaching @ColumbiaSIPA | Affiliate @BKCHarvard @Graphika_NYC | Former @Google
Jan 28, 2022 • 13 tweets • 10 min read
🆕 📖 Here is a short 🧵 about my latest publication, a research report published with @AJUnited looking at the applicability of bug bounties to the discovery and redress of algorithmic harms!
It’s such a joy to see a project you’ve worked on for 2+ years finally see the light of day ☀️.
This work started in Nov. 2019 during a Bellagio residency where @JovialJoy and I started exploring bug bounties for the participatory discovery and redress of algorithmic harms.
Mar 8, 2021 • 12 tweets • 5 min read
A few folks have asked what I covered in this intro to attributing information operations.
Threaded below is a high level recap of the structure of our discussion. Planning to share full details in a writeup, as some of the details and examples matter here.
Attribution here is “simply” the practice of identifying who are the entities responsible for conducting disruptive, covert operations. As such, that question only applies to a very small section of the universe of mis/disinformation campaigns - worth keeping in mind!
Feb 7, 2021 • 12 tweets • 3 min read
I've just been down a minor rabbit hole trying to understand why US-bought "Herbes de Provence" tasted so weird for me and made my croque monsieur taste off. The answer was both confusing and lovely.
I realize this is a very French thread, my apologies in advance.
As many others, the pandemic and closed borders means I've been separated from my family for almost a year. I'm homesick and clinging to whatever can bring a taste of home in NY. This has resulted in an increase in French cooking at home despite my being quite the incapable cook.
Jun 16, 2020 • 21 tweets • 13 min read
🆕 Today, we are exposing 6 years of "Secondary Infektion" activity, a persistent Russian information operation network which we've been investigating for months.
It lives here, in 150 pages of details, with its database and carousel of forgeries (!):
secondaryinfektion.org
Secondary Infektion:
📌 Spans at least 2014-2020
📌 Loves forgeries
📌 Operates in at least 7 languages
📌 Maintains high OPSEC throughout the years
📌 Leverages inauthentic single-use accounts on multiple platforms
📌 Achieved minimal traction online (besides UK op., Nov. 19)
Mar 2, 2020 • 7 tweets • 2 min read
📑We wrote about a campaign run by a PR firm in India which focused on geopolitical rivalries in the Gulf, often through the lens of professional football (this: ⚽, not that: 🏈) . Thread below👇and full report here: graphika.com/reports/operat…
This campaign highlights a few recent trends in influence operations.
Oct 1, 2019 • 8 tweets • 4 min read
Last week, I headed to Congress to testify before the House Committee on Science, Space and Technology. I was grateful for the opportunity to share a few thoughts with Representatives for their hearing on “online imposters and disinformation.” Select highlights in this thread👇
Too often, disinfo is portrayed mostly as a “content” problem (ex. fake news, deepfake). Content is an important dimension, but I argued that the “actors” behind the campaigns and the “behaviors” employed to manipulate audiences are underappreciated yet crucial vectors in disinfo
