AI safety research has been concentrated in a handful of wealthy countries, leaving critical blind spots in how AI systems behave in different languages, cultures, and economic contexts. This weekend, a new Global South AI Safety Hackathon is attempting to change that by bringing together researchers, engineers, and policy professionals from Latin America, Africa, and Asia to tackle AI problems that matter most in their own regions.

Why Is AI Safety Research So Concentrated in the West?

The numbers tell a stark story. None of the top 100 institutions by AI publication index, in either universities or companies, are based in Africa or Latin America. This concentration means that the vast majority of AI safety work reflects the priorities, languages, and regulatory landscapes of wealthy nations, leaving significant gaps in understanding how AI systems fail or cause harm in other parts of the world.

The consequences are real. Jailbreaks, which are attempts to manipulate AI systems into ignoring their safety guidelines, are more common in low-resource languages. Algorithmic bias trained on non-local data shows up in healthcare and hiring systems deployed in regions where the training data never reflected local populations. These are not theoretical problems; they affect real people's access to credit, medical care, and employment.

What Problems Are Researchers in the Global South Trying to Solve?

The hackathon, running from June 19 to June 21, 2026, is structured around three regional tracks, each with its own focus areas and local context.

• Latin America: Participants are working on AI fairness for Portuguese and Spanish language models, safety evaluations for systems deployed in Latin American contexts, and regulatory analysis of emerging legislation. Brazil's AI bill includes a standalone human rights chapter that goes beyond the European Union's AI Act, while Chile became the first country in the world to constitutionally protect neuro-rights.
• Africa: The focus includes deepfake-driven electoral interference, data-colonial dependency on foreign infrastructure, compute and semiconductor scarcity that constrains sovereign AI capacity, and large-scale labor market disruptions. Teams are building solutions for misuse resilience and differential defense acceleration.
• Asia: With India's "Seven Sutras" framework prioritizing innovation, China's sector-specific regulations, and Vietnam's new AI law that took effect in March 2026, the region spans the full spectrum of governance approaches. Projects address cross-border governance harmonization, safety evaluations for non-English language models, and region-specific risk assessments.

The hackathon is deliberately designed to bring the Global South into AI safety research, not to bring AI safety to the Global South. Researchers in these regions have contextual knowledge the field desperately needs: understanding of regulatory landscapes, language gaps, and institutional constraints that determine whether safety research actually works in practice.

How Can Researchers and Engineers Get Involved?

The hackathon welcomes a broad range of participants and skill levels. Teams can form around specific problems and choose from regional tracks and sub-tracks tailored to their context.

• Participant Types: The hackathon is open to AI safety researchers and engineers, machine learning researchers, policy researchers working on AI governance, software engineers interested in safety infrastructure, security researchers and red teamers, students and early-career researchers, and anyone working on AI and its impacts in the Global South.
• Project Formats: Teams can build tools, evaluations, policy analyses, or research contributions. They research and scope a specific problem, build their project over the weekend, and submit a research report documenting their approach, results, and implications.
• Regional Hubs: Latin America has hubs in São Paulo, Buenos Aires, Bogotá, and Mérida-Guadalajara. Africa is represented through Cape Town. Asia includes Bengaluru, New Delhi, Hanoi, and Ho Chi Minh City. Participants compete within their region, not globally.

The top teams from each region receive prizes, with Latin America offering 3,000 dollars total and Asia offering 2,000 dollars total. More importantly, the best teams are invited into the Apart Fellowship for continued research and mentorship.

Why Does This Matter for Global AI Governance?

The fragmentation of AI regulation worldwide is creating new challenges for businesses and policymakers. Different jurisdictions define "AI" differently, use different legal forms for regulation, and apply different conceptual approaches to enforcement. Some regulations are legally binding while others are merely guidelines. Some are sector-specific while others apply across all sectors. This inconsistency means that international businesses often face a "highest common denominator" approach, adopting the strictest applicable standard to ensure compliance everywhere.

By building AI safety research capacity in the Global South, the hackathon addresses a critical gap. When researchers from Latin America, Africa, and Asia contribute to the global conversation about AI safety, they bring evidence about how AI systems behave in their contexts. This knowledge can inform better regulatory frameworks that account for the diversity of AI deployment scenarios worldwide, rather than frameworks designed primarily around Western use cases.

The timing is significant. Vietnam's AI law took effect in March 2026, making it the first binding AI legislation in Southeast Asia. Brazil's AI bill is pending a Chamber vote. Colombia established a national AI policy framework in early 2025. These regulatory developments show that the Global South is not waiting for Western guidance; it is building its own frameworks. Research from this hackathon can directly inform these efforts.

The broader message is clear: AI safety is not a problem that can be solved by researchers in a handful of wealthy countries. The risks are global, the deployment contexts are diverse, and the solutions need to reflect that reality. This hackathon is a step toward building the research infrastructure and talent pipeline that makes that possible.