EU AI Act, DMA, and AI Regulation: What Enterprises in Latin America Need to Know
The EU AI Act is in force and the DMA has already required Google to open Android to rival AI assistants. Analysis of how these regulations affect Latin American companies that use AI, export software to Europe, or have European clients.
Key Takeaways
- The EU AI Act is in force: high-risk AI systems require certification and audits before operating in Europe.
- The DMA required Google to open Android to rival AI assistants (Claude, ChatGPT) before August 2027.
- Mandatory watermarks on AI-generated content are now required for models launched from August 2026.
- Latin American companies that sell software to European clients or handle European citizen data are subject to the EU AI Act.
- 'EU AI Act compliance' will become a procurement requirement in global enterprise tenders within the next 12-24 months.
The EU AI regulatory framework has moved beyond a theoretical draft: the EU AI Act is in force with staggered compliance deadlines, the Digital Markets Act (DMA) has already issued binding orders to Google, and mandatory watermarks on AI-generated content are now a standard for certified models.
For Latin American companies, this is not exclusively a European problem. EU regulations have extraterritorial reach: any company that processes data from European citizens, sells software to EU clients, or has operations in European markets is subject to these regulations. Ignoring them can result in exclusion from enterprise tenders, significant fines, or barriers to European market access.
In July 2026, the EU issued binding orders under the DMA requiring Google to open 11 features of its Android operating system to rival AI assistants like Claude and ChatGPT before August 2027. This decision has implications for any company building mobile applications on Android that wants to offer integration with multiple AI assistants.
EU AI Act: Risk Classification and Obligations
The EU AI Act classifies AI systems into four risk categories. Unacceptable risk systems (subliminal manipulation, mass social scoring) are prohibited. High-risk systems (AI in critical infrastructure, employment, education, essential services) require conformity assessments, extensive technical documentation, and registration in a European database. Limited-risk systems (chatbots, deepfakes) require transparency toward users. And minimal-risk systems (spam filters, AI games) have no additional obligations. For most companies, the critical analysis is determining whether their systems fall in the 'high-risk' category.
The DMA and the AI Assistant Market
The DMA order to Google has a 'level playing field' effect on the mobile AI assistant market. Before August 2027, Android users in Europe will be able to choose which AI assistant to use as default, with equivalent access to system APIs. For companies developing mobile applications, this creates an opportunity: designing provider-agnostic AI assistant integrations will be a competitive advantage. Architectures that hardcode integrations with a single AI assistant will be harder to adapt.
Practical Impact for Latin American Enterprises
If your company exports SaaS software to European clients, manages data from European employees or clients, or aspires to expand operations to Europe in the coming years, the EU AI Act is already relevant to you today. Recommended immediate actions: 1) Inventory all AI systems in use or development and classify them according to the EU AI Act risk scheme. 2) Review AI provider contracts to confirm they offer compliance guarantees. 3) Implement transparency policies about AI use in products toward end users. 4) Document datasets used to train or fine-tune proprietary models. Proactive compliance today avoids costly disruptions tomorrow.
AI Governance as Competitive Advantage
Companies that develop solid AI governance practices today will not only avoid regulatory problems: they will have a procurement advantage in global enterprise tenders. In the next 12-24 months, 'EU AI Act compliance' will become a checkbox in vendor evaluation questionnaires from Fortune 500 companies and governments. Companies that can demonstrate audits of their AI systems, risk documentation, and transparency controls will have preferential access to those contracts. At Nubework, we help our clients build these governance frameworks from the design of their systems.
Sources
Want to apply this in your organization?
Talk to a Nubework architect.
