Community
A community makes AI safety work sustainable: people meet, collaborate, and stop advancing on their own.
About Us
A Colombian chapter focused on the risks of advanced AI systems. Here we share who we are, how the group started, and the three levers that shape our work: community, learning, and events.
Mission
Reduce the existential and catastrophic risks of artificial intelligence.
We develop local talent, produce relevant research, and connect Colombia with the global AI safety community.
Why now
Why this matters
Artificial intelligence is advancing faster than our ability to understand and govern it. These are the ideas that guide our work.
-
AI is outpacing our institutions
AI systems are improving faster than laws, universities, and governments can adapt. That gap, between what the technology can already do and what we are prepared to handle, is where the greatest risks emerge.
-
The right people, early, change the trajectory
Every technology is shaped by who is in the room when the decisions get made. Training researchers and policymakers in AI safety today is the most concrete way to influence how it unfolds.
-
The Global South can't be left out of this conversation
AI decisions are concentrated in a handful of Global North centers, but their consequences belong to everyone. Colombia and Latin America need their own voice and talent in how this technology is built and governed.
Our story
How we started
AI Safety Colombia started in 2024. Its first public activity was a talk on the ethical dilemmas of superintelligence, at Universidad de los Andes, in September of that year. The organization was cofounded by Jose Gelves and Fernando Avalos-López.
That same year we launched our first two fellowships, in AI governance and technical AI safety, with weekly sessions and one-on-one mentorship. The first ML4G in Colombia, a fellowship on the economics of AGI, and discussion groups on AI policy followed. Alongside them runs a steady rhythm of talks, panels, and dinners that bring the community into contact with the country’s AI researchers and policy figures, among them authors of Colombia’s AI ethical framework and Global Partnership on AI members.
Today we coordinate the Bogotá hub of Apart Research's Global South AI Safety Hackathon, connecting the local community with an international AI safety research network. Our field-building work has been supported by a grant from Coefficient Giving.
- 2024
First public talk, on superintelligence, at Universidad de los Andes.
- 2024
The community's first two fellowships: AI governance and technical AI safety.
- 2025
First ML4G in Colombia, with around 20 participants.
- 2025
A fellowship on the economics of AGI and the first AI policy discussion group.
- 2026
Bogotá hub of the Global South AI Safety Hackathon.
Theory of Change
How we think change happens: three levers that reinforce each other.
Learning
With solid training, more people move from interest to real contribution, in technical research or AI governance.
Events
Gatherings connect the community with researchers and policy figures, and open doors to the global field.
Where we're heading
What we're building
Those three levers point toward three long-term goals, the ones that set the direction for AI Safety Colombia.
Research from the region
We aim for original AI safety research to come out of Colombia and Latin America. Today we are taking the first steps: connecting the community with the training, tools, and opportunities to get there.
A growing community
We want more people in Colombia, students, professionals, and researchers, to find in AI safety a field where they can contribute.
A voice in the public conversation
We want AI safety to have a place in the policy and technology debate in Colombia, bringing rigorous information to decision-makers.
Team
AI Safety Colombia is a small, volunteer-led organization, sustained by a wider community.
In memoriam
Fernando Avalos-López
Cofounder
AI alignment researcher, Apart Research
Systems engineer from Universidad de los Andes. He was among the first to advance AI safety research in the country.
Fernando helped imagine and build this community from the start. His memory remains present in everything we do.
Jose Gelves
Cofounder & organizer
AI policy and digital-government researcher
A political scientist from Universidad de los Andes, interested in how technology gets governed. He has researched AI in elections and employment, evaluated Colombia's public-sector digital transformation, and leads Effective Altruism Bogotá.
Leonardo Párraga
Organizer
Coalition-builder across civil society, philanthropy, and AI governance
He holds an Ed.M. in International Education Policy from Harvard and has worked across 25+ countries. He founded BogotArt and the Colombian Coalition on Youth, Peace and Security, and channels resources toward the highest-impact causes in Latin America.
Be part of this
This community is built by people. There is a place for you, wherever you are starting from.