We're a civic technology lab working to strengthen democratic processes by creating digital tools that support free and fair elections, open governance, and fact-based public policy.

What We Do

We use the latest advancements in distributed computing and machine learning to detect and counter malicious information campaigns.

We are coders who are deeply concerned about the state of transparency, accountability, and fact-based policy in our democracy. So we decided to do what we know best — using technology in order to effect change. We won’t let the malicious actors win the arms race.

That’s why we’re using cutting-edge techniques in order to call out bad actors and stop falsehoods from spreading so that healthy debate can happen. We know that once these democratic norms disintegrate, they’ll disappear from our public conversation forever.

Detecting Campaigns

Our crawlers search the digital world for six indicators of malicious narratives: messages amplified through automated means, messages that fail fact checks, messages originating from false profiles, messages advocating for violence against specific people or groups, messages that have unnatural viral curves, and messages that correlate with known malicious narratives.

Countering Narratives

Once we understand which narratives are part of malicious information campaigns and how these messages are being distributed, we then carry out operations to reduce their saliency by pushing relevant information to stakeholders and pressuring them to take action.

How We Do It

We investigate trending narratives and reverse engineer distribution channels with a combination of signal disruption and targeted counter-messaging.
Evaluate Trending Narratives

We start by crawling the internet for narratives that exhibit two or more of our six indicators of maliciousness.

Track Propagation Timeline

Next, we keep an eye on the propagation timeline in order to prioritize narratives that are ripe for disruption.

Map Distribution Channels

We then break down the flagged narratives by distribution channel to understand how the messages are disseminated.

Measure Engagement

We also keep an eye on the engagement that narratives receive so that we can isolate the most impactful campaigns.

Build Audience Segments

In addition, we look to see if specific demographics or interest groups seem to have been targeted by the flagged narratives.

Deploy Countermeasures

Finally, we execute both defensive and offensive operations in order to stamp out the malicious information campaigns.