Catalyze talent
Mentor and evaluate the next generation of AI safety researchers, engineers, and policy professionals from underrepresented regions.
The Global South AIS Hackathon
An international hackathon convened by Apart Research and AI Safety Mexico. It brings together researchers, builders, and policy professionals from Latin America, Africa, and Asia to produce concrete work in AI safety, anchored in regional context.
Registration is open for participants and collaborators.
The mission
What is this hackathon?
An intensive weekend to develop prototypes, evaluations, and governance proposals in AI safety, connecting the Global South community with international researchers and mentors.
Organized by Apart Research and AI Safety Mexico as the regional partner, this initiative decentralizes AI governance and technical research, elevating local expertise onto the global stage.
AI safety cannot be solved from a single region. The Global South brings irreplaceable context, expertise, and perspectives to this conversation.
Why it matters
Why collaborate?
Four reasons this hackathon is relevant to those working on AI safety and governance.
Solve real problems
Support solutions grounded in regional challenges: AI safety, governance, and accountability across the Global South.
Decentralize the conversation
Help move the global AI safety agenda beyond a handful of institutions, toward a genuinely international ecosystem.
Build visibility
Connect with researchers, founders, and policymakers across three continents and gain presence in a growing international initiative.
Talks and workshops
Speakers
Voices joining the hackathon with talks and workshops on AI safety, governance, and data sovereignty.
Damiano Fornasiere
LawZero
From mathematics to technical AI safety
Shares his path into AI safety and describes his current work at LawZero.
Luis Cosio
SL5 Task Force
How to build AI that doesn't spiral out of control
An introduction to AI alignment and why Latin America matters to the field.
Nikita Lokhmachev
Greyhaven
Your AI Agent Is Not You: AI safety, data sovereignty, and taking back control
Your AI agent runs with your permissions, credentials, and file access, guided by a model you didn't train. A talk on AI safety and data sovereignty, with a hands-on demo wrapping Claude Code with Greywall.
Camilla Balbis
AIGS Canada · SPAR Research Fellow
The Governance Copy-Paste Trap
How the Global South can learn from the North's missteps on AI safety and governance, and why grassroots, self-determined domestic resilience (not imposed frameworks) is a better path forward.
Marcos Galván López
Instituto Politécnico Nacional
Prior knowledge as a philosophy to design safe intelligent systems
How incorporating prior knowledge about the problems we aim to solve helps design AI systems that are safer, more reliable, and auditable.
Fernando Castillo
Threat Trackers
AI as a cognitive weapon: manipulation, propaganda, and national security in the age of the synthetic society
How AI turns disinformation into an automated, scalable, personalized system (from bots to deepfakes) and why cognitive defense (critical literacy, algorithmic transparency, responsible regulation) is key to democratic resilience.
Panel
Judges
Those who will evaluate the projects and give expert feedback to participating teams.
Cibeles García Burt
USAID
Mérida, Mexico
U.S. diplomat with 17 years of experience across Latin America, Africa, and Asia. Now focused on AI governance, with several BlueDot courses completed.
Camilla Balbis
AIGS Canada · SPAR Research Fellow
Toronto, Canada
Multilingual AI governance policy analyst. Researches with AIGS, combining risk analysis, technology ethics, and international collaboration to strengthen the responsible development of AI.
Luis Cosio
SL5 Task Force · Member of Technical Staff
Guadalajara, Mexico
Works at the intersection of frontier AI and high-security systems, turning abstract safety requirements into deployable solutions resilient to real adversaries, including nation-state attacks and loss-of-control scenarios. Winner of multiple Apart hackathons.
Nikita Lokhmachev
Greyhaven · Technical Product Lead
Mexico
Senior ML engineer and co-founder with a master's in AI. After Apart hackathons, he shifted his career toward AI safety.
Vincent Mai
LawZero · Senior Research Scientist
Montreal, Canada
PhD in Machine Learning from Mila.
Jesús M. Siqueiros
IIMAS, UNAM
Mérida, Mexico
PhD in Philosophy of Science. He researches participatory modeling, sustainability, and AI. In 2024 he received the Google Academic Research Award for Community Weavers of AI and Nature-based Solutions and leads the Lab-DEMOSS project.
Fernando Castillo
Threat Trackers
Socio-technical cybersecurity researcher working at the intersection of digital security, human rights, democracy, and AI safety. Founder of Threat Trackers, a civil-society organization democratizing cybersecurity for underserved communities.
Marcos Galván López
Instituto Politécnico Nacional
Researcher in neurosymbolic AI, explainable machine learning, and formal and probabilistic reasoning. He integrates neural networks with symbolic methods to build models that are more interpretable, auditable, and reliable. He has conducted research at the University of Edinburgh.
Amanda Isbosseth Guzmán Aceves
Government of Jalisco
Nanotechnology engineer from UdeG. Former Vice President of IEEE CU Tonalá and former Chair of Women in Engineering Guadalajara. Fulbright alumna at UT Dallas and former research intern at ExxonMobil Houston. Winner of ANUIES4MX 2025 with a research stay at ATR Kyoto. She coordinates data science initiatives at Jalisco's Secretaría de Administración.
Amaia Amézaga
Sattva Labs · Founder & Independent Researcher
Independent researcher and founder of Sattva Labs, focused on AI safety, model evaluation, and human-AI interaction. She combines interaction design (HCI), multicultural communication, and qualitative analysis to study how people relate to AI systems.
Collaboration roles
How you can contribute
There are multiple ways to join the event, from one-off mentoring to institutional sponsorship.
Judges
Evaluate projects and provide expert feedback. Ideal for researchers, engineers, and policy professionals with AI safety experience.
Mentors
Share knowledge on AI responsibility, safety, and governance in the Global South across the weekend.
Speakers
Deliver short talks or workshops on AI safety topics. Broaden the conceptual and practical toolkit of participants.
Volunteers
Support coordination, logistics, and participant experience at the in-person hubs (Mérida and Guadalajara) and in the remote format.
Sponsors
Fund prizes, venues, or ecosystem development. Gain prominent visibility in an international AI safety initiative.
Event snapshot
Save the date
Three essentials for planning your participation.
- Format
- Primarily remote, with in-person hubs in Mérida and Guadalajara, Mexico.
- Dates
- June 19–21, 2026. Three intensive days in parallel across Latin America, Africa, and Asia.
- Prizes
- 3,000 USD prize pool: three winning teams in Latin America will each receive 1,000 USD.
Sponsors
With support from
Institutions and partners making the hackathon possible.
Apart Fellowship
Beyond the hackathon
The top teams from every global hub will be invited to the Apart Fellowship: a pipeline of emerging AI safety talent with continuing mentorship, research opportunities, and professional support.
Support the future of AI safety in the Global South
Join us in building safer, more contextualized AI systems by supporting the next generation of researchers, builders, and policymakers across three continents.
Get in touch to collaborate