Revontulet

What toxic narratives shape the online conversation in and about Malm? on TikTok and X? ? We continue our pursuit to get new insights on polarization on social media. ? Together with Malm? stad and Revontulet, we have analyzed over 10,000 videos and tweets from June 2023 to May 2024?to understand which toxic narratives are dominant, how they move across the platform narratives and what methods are used to amplify the narratives.?Here’s what we found: ?? Different toxic narratives in different languages:?In Swedish, the dominant toxic narratives was the Great Replacement?theory, fears of Islamization, and criticism of the Social Democrats for creating a “haven for terrorists.” In Arabic, the dominat toxic narratives was on Islamophobia in Sweden, with recurring themes of Koran burnings, child abduction (LVU-campaign), and the Israeli-Palestinian conflict. ?? Online and offline behaviors are interconnected: Real-world events shape online discussions, fueling toxic narratives. At the same time, online discourse can influence actions in the physical world. ?? Superspreaders drive polarization: A small group of influential individuals and organizations play a key role in amplifying toxic narratives, shaping public debate on TikTok and X. Three years ago, we launched the?Safe Digital City?initiative in Malm?. Since then, the city has been at the forefront of digital security efforts now with analysis of facebook, Tiktok, Flashback and X. ? This report marks the next step—now, the crucial work of translating insights into action begins. Listen to the deputy mayor of Malm? stad Amani Loubani talk about the analysis on Sveriges Radio: https://lnkd.in/dcxdQW38 Read more about the analysis here: https://lnkd.in/dG_2zMGQ #NordicSafeCities #SafeDigitalCity #democracy #toxic #narratives #polarization

This weekend, Revontulet will attend the Paris AI Action Summit. The summit will gather Heads of State and Government, leaders of international organizations, and CEOs of small and large companies to discuss the future of AI. Representatives of academia, non-governmental organizations, artists, and members of civil society will also attend. On Sunday, February 9th, we are excited to contribute to critical conversations about AI ethics and responsible technology with our partner, Humane Intelligence, at the AI & Society House. Revontulet will participate in panel discussions and have an expo booth. We will present our previous work with HumaneIntelligence's Bias Bounty to uncover visual extremist propaganda. We will also discuss our overarching work, thinking, and objectives concerning the evolving threats and opportunities AI offers. Revontulet provides risk analysis, open-source intelligence (OSINT), and data services on terrorist and violent extremist networks and their behaviors to clients in the tech sector and beyond. While we offer conventional intelligence, the gathering, and analysis of information developed to inform actions to prevent harm, we also consider the term and its meaning more broadly in this era of Artificial Intelligence. Read more about our approach to AI and our participation at the Paris AI Action Summit in the article linked in the comments.

We look forward to joining our friends and partners at HumaneIntelligence for the AI & Society House during the Paris AI Action Summit on February 9th. As AI plays an increasingly large role in the Terroist and Violent Extremist Content (TVEC) landscape, possessing the power to inspire real world harm and spreading toxic, polarizing and conspiratorial online content at an unprecedented scale we are excited to share more about both the impact we see of AI in our field, and how we apply artificial intelligence to conduct research, advance our field, improve our processes and create a safer world for all.

We were excited to partner with our friends at HumaneIntelligence last month for one of their bias bounties, tackling some of the challenges posed at the intersection of AI and ethics in our field of counter terrorism and the spread of hate-based propaganda on social media. Through the challenge contestants developed AI solutions capable of detecting and interpreting manipulated images used by extremist groups. In Humane Intelligence gathered some of the feedback from some of the participants, key takeaways we will carry with us as we continue to tackle hate-based propaganda, violent extremism and terrorist abuses of our shared tools and infrastructure. Here's some of the feedback. From Mayowa Osibodu: “This hackathon was an exciting learning experience. Tackling hate-based propaganda through AI solutions is vital work, and it’s inspiring to have organizations like Humane Intelligence and Revontulet bringing people together to develop impactful counterterrorism tools. The challenge allowed me to explore new approaches and sharpen my skills in ways that will shape my future work in AI and beyond.” From TUES DAY/Artifex Labs: “This challenge was an incredible opportunity for me to merge my patriotism with my passion for data science and AI ethics.[.] Winning this challenge has validated that decision and reignited my commitment to serving my nation through innovative work like this. As a Latina of Indigenous heritage, I’ve faced the absence of mentors in this field. I taught myself the skills I needed—from machine learning to Python—using online resources and academic papers. Programs like Bias Bounty show that with dedication, anyone can rise to meet these challenges, regardless of academic pedigree, background, or personal hurdles. From T. Devon Artis: “The Bias Bounty program was a game-changer! I jumped at the chance to tackle AI bias head-on and be part of something so groundbreaking. The incredible support from Nicole and the Humane Intelligence team made this challenge more than just a competition—it was a pioneering learning experience. Finding and addressing AI bias is critical work, and this program proved how innovative solutions can have a real-world impact. I’m already looking forward to the next round and can’t wait to see what’s ahead.” Find the full article from Humane Intelligence in the top comment, and check out their upcoming challenges. We're thrilled to have partnered with Humane Intelligence for this challenge, and look forward to many opportunities to work with them and their community in the near future.

Congratulations to all the winners of the Bias Bounty! We partnered with HumaneIntelligence to challenge leading AI researchers in developing computer vision models to detect extremist propaganda images. We look forward to digging into all the solutions offered, and see how they advance the use of AI across our field.

?? Reminder: Today is the final day to submit for Bias Bounty 2! If you haven't sent in your submission, you have until 11:59 PM ET tonight to be part of Humane Intelligence’s second algorithmic bias bounty challenge, in partnership with Revontulet. ?? Our mission: Use AI to identify hidden extremist propaganda and advance counterterrorism efforts. This challenge focuses on building a CV model to uncover and interpret manipulated, hateful content, helping to reveal the strengths and gaps in machine learning’s role in keeping communities safe. ?? Prizes totaling $10,000: $4,000 for intermediate winners $6,000 for advanced winners ?? Don’t miss out! Submit by tonight to make your mark in this critical field.

Our Data & Research lead Chardon Murray was interviewed by WHQR Public Radio about the sudden appearance of election posters from the far right group Patriot Front across Wilmington, North Carolina ahead of the U.S. Election. Check out the interview and insights she have to share here: https://lnkd.in/dgBS-ibf

No day shall erase you from the memory of time. Today I was honored to be shown the 9/11 memorial in NYC by survivors who shared their experiences of that day, and its aftermath with myself and the Revontulet team. I will forever be grateful for that experience. As a survivor-founded company we never forget why we do what we do. Still, it is important for us all to, from time to time, stop for a moment to acknowledge it. This week I was fortunate enough to bring together some of our team as we visited Eradicate Hate Global Summit and New York. Both provided opportunities to connect with fellow survivors, to hear their stories and recognize their struggles. On our way from Pittsburgh to NYC we also stopped by the memorial to Flight 93. While we every day live and breathe our mission to end terror to make the world safer for all it is important that we remember those who are no longer with us, those who live on and that they inform why we do what we do. Countering terrorism is not an abstract mission, an opportunity to advance careers, or an intellectual exercise. It has real world impact, and it matters. Our work cannot, and should never be detached from the reality of terror and the devastation it causes. We must always remember why we do what we do. As a founder, colleague and survivor I’ve been honored to share these days with our amazing team, and with colleagues from across the sector. Our experiences with terror are varied, but coming together, listening and acknowledging it’s human impact is vital. It makes it real. I am grateful to my colleagues, and to my fellow members of “the club nobody should ever want to join,” as we shoulder by shoulder continue to strive for a safer world for all.

We've got our table set up at the Eradicate Hate Global Summit featuring an overview of Revontulet's network maps, services, and projects! Having so many awesome conversations. So excited to share the fruits of our labor!

?? Reminder: Sign up for Bias Bounty 2! ?? ?? Don’t miss your chance to participate in our second Bias Bounty challenge, in partnership with Revontulet! After a successful first edition, we're back with a new focus: counterterrorism in computer vision (CV). ?????? Help us develop models to identify far-right extremist content disguised on social media, and uncover gaps in the use of AI as a counterterrorism tool. ?? Intermediate and Advanced levels are available, so there's something for everyone ready to make a real impact in AI safety and security! ?? Ready to take on the challenge? Our Fellow Ay?egül Güzel tells you more on this video. Sign up using the link in the comments.