This article is written by Enrique Zapata, Principal Executive for Data Intelligence and New Technologies in CAF, and Constanza Gomez-Mont, CEO of C-Minds

Artificial Intelligence (AI), once the stuff of sci-fi and Hollywood movies, has become part of daily life.

Today, AI is used in areas like mobility, retail, hospitality and will increasingly be used in social programs, security and finance. In the pandemic we are currently in, AI and the data that underpins it will be key for healthcare policy, vaccine research and risk management. Eventually, in the next 5 to 10 years, virtually all areas of our life — either directly, or indirectly — will be influenced by AI.

While we have a good idea about where the field of AI is headed, there are many relevant questions that remain unanswered: What data will these applications use? Where will this data come from and what are their biases? How should the big tech companies like Google and Facebook treat personal data, inside and across countries? How will jurisdictions address ethical issues around algorithmic accountability, openness and transparency? How should we protect human rights, while at the same time unlocking an economic potential of close to 16 trillion dollars by 2030?

These are only some of the questions that make the need for national and local AI strategies relevant, since the path governments decide to follow today will define their place in an increasingly AI driven global space.

In the next 15 years, 19% (9.8 million) of all jobs in Mexico will potentially be exposed to automation

Everything points to the existence of more than one model. In parts of Asia, states control how these technologies are being designed and deployed. In North America, a laissez-faire approach focuses on how the private sector develops and uses AI, with minimum involvement from the state.

In Europe a mixed approach is taking shape, focused mainly on a strong regulatory framework enforced by the state, and implemented through big techs and SMEs. In Latin America, Africa and Southeast Asia it’s still an open question how this will play out. Countries in these regions have to decide whether they will lead, or will fall into an umbrella of others, and in that case if they are forced to follow with no or minimum voice on how AI will transform their governments and societies.

Mexico and AI

For these reasons the Government of Mexico decided in 2017 to kickstart a national conversation about the future of AI. The goal was to decide which of the international examples Mexico should follow through three concrete actions: first, by analysing the country’s readiness for AI; second, by launching its first AI Strategy in 2018, becoming one of the first 10 countries to do so; third, by consolidating the AI multi-sector ecosystem.

First action: Readiness assessment and policy recommendations

In the next 15 years, 19% (9.8 million) of all jobs in Mexico will potentially be exposed to automation

To understand the readiness of Mexico for AI, the Governments of Mexico and the United Kingdom (UK) commissioned a study and Policy Report through the UK’s Prosperity Fund done in collaboration with AI leading organisations C-Minds and Oxford Insights. The report, titled “Towards an Artificial Intelligence Strategy in Mexico: Harnessing the AI Revolution”, analysed the opportunities and challenges of Mexico and summarising the perspective of more than 80 national experts in the AI field.

This report which laid the foundation for the later AI strategy, focuses its policy recommendations for the short, medium and long terms in 5 main areas:

  1. Governance, government and public services, by using the transformative power of AI for public service delivery, in spheres such as health, education and social care, as well as for improving the inner workings of government, increasing public sector efficiency and even assisting in the policy-making process.
  2. Research and development, by providing support for building AI applied knowledge through government and catalyse markets to do likewise.
  3. Capacity, skills and education, by introducing coding, tech skills, and science, maths and engineering early in the national curriculum, support AI fellowships at postgraduate level and lifelong education and training to ensure the workforce can adapt to changing conditions.
  4. Data infrastructure, by supporting the release of reusable data from government and public interest data from the private sector in formats which AI can make sense of, and by supporting high-quality digital infrastructure such as broadband networks and 5G.
  5. Ethics and regulation, by implementing guidelines and codes of practice for AI as ways of addressing ethical issues.

By publishing this report Mexico’s government created the foundation for a better understanding of capacities and needed actions to move forward.

Second action: Developing Mexico’s AI Strategy

With the assessment and recommendations of the report in place, the Office of the President, in collaboration with civil society and academia, launched the first AI strategy for the country and Latin America in March 2018.

Mexico's AI Strategy 2018, included five key actions:

  1. Develop an inclusive governance framework, through the creation of an AI Sub-commission led by the Office of the President and with official participation of industry, civil society, academia and some local governments.
  2. Identify the uses and needs of AI in industry, through a documentation exercise carried out by the Mexican Chamber of Telecommunication Industry, as well as identifying best practices within government of AI.
  3. Open the medium and long term recommendations of the Policy Report for public consultation;
  4. Support Mexico’s AI leadership in international forums, including the OECD, D9 and G20; and,
  5. Promote continuity through changing administrations, by working with all interested stakeholders towards an official AI National Policy.

The government had also published two key documents related to responsible and ethical development and use of AI: AI General Principles and Risk Assessment tool for the development and use of autonomous systems in government. On the international stage, Mexico took the initiative to form a UN-working group that had the participation of 34 countries; and lead  the Working Group for Emerging Technologies in the Latin America and Caribbean Network for the Development of Digital Governments (GEALC, its acronym in Spanish).

Together, these actions made it possible to build the legal bases for AI governance mechanisms in the public sector, while at the same time creating ownership and positive incentives in other sectors to promote the sustainability of the strategy in the long run.

Third action: Consolidating and Engaging the AI ecosystem in a sustainable way

Following the recommendations from the Policy Report, the government launched a multi-stakeholder national coalition in 2018 in partnership with institutions from the industry, civil society, and academia.

The goal of this coalition — IA2030.mx — is to coordinate efforts, build a single voice around AI and promote continued action and support for the topic. What started as the collaboration of 10 institutions grew to the participation of more than 110 on a national scale.

Mexico offers a blueprint for other governments that want to develop their own AI strategy

From this broader understanding of diverse perspectives, the quest for developing bottom-up recommendations for a National AI strategy started by the end of 2019. Chaired by the co-founding organisation, C Minds, six working groups were established, each with their own specific focus: 1) AI Ethics 2) Governance, Government and Public Services 3) Research and Development, 4) Skills, Capacities and Education 5) Data, Digital Infrastructure and Cyber Security, and 6) Mexicans abroad.

Today the coalition is a key participant in diverse AI related events with a single voice, and through its members it is currently participating in legislative initiatives around the future of AI in Mexico.

Moreover, the creation of this coalition has consolidated the AI ecosystem, bringing together actors from all sectors working in the AI field to find common grounds that allow for a joint effort to strengthen Mexico's capacity to harness the power of AI while mitigating its social and environmental risks.

Conclusion and recommendations

As we’ve seen with Mexico’s case, the role of governments in the creation of AI strategies and national policies is crucial, since they can set the direction and solid and ethical foundations for AI development.

Following national elections In the spring of 2018, Mexico had a change of administration, and much of the work from the government around digital took a halt with a shift in policy priorities. And while some may see this as a set-back, it shouldn’t overshadow the progress that has taken place: Mexico is the first country in Latin America to develop an AI strategy, as well as one of the few countries in the world with an official roadmap on how to develop and adopt AI. Furthermore, the work of the coalition ia2030.mx is stronger than ever, as it continues to consolidate the AI ecosystem in Mexico from civil society, private sector and academia, and is collaboratively exploring opportunities for the country.

Mexico offers a blueprint for other governments that want to develop their own AI strategy. Here are five key recommendations based on Mexico’s work in this field:

  1. An initial benchmark of where a government and ecosystem stands with relation to AI is needed to make informed decisions and develop an action plan.
  2. Strategic, multi-stakeholder partnerships are needed to align efforts across different sectors and to make sure that open participation channels are established so that the developed AI strategies are inclusive and integrate an array of diverse perspectives
  3. Involve other branches and levels of government, to create and maintain inertia, support sustainability and enable the design of comprehensive national AI agendas.
  4. Keeping human rights and distributed benefits at the centre of the AI strategy is key, especially to make sure that the development and use of AI enables shared benefits and avoids widening existing social gaps.
  5. Public and private funding is critical, to achieve robust and sustainable AI strategies aimed at transforming a national environment in the long run.

Mexico, like most of the world, still has a long way ahead to consolidating efforts towards the harnessing of AI for the benefit of its populations and economy. At the core of these efforts, there should be a recognition that AI strategies are needed to enable a swift transition into the AI driven digital economy, leverage AI for increased competitiveness and productivity and most importantly, to protect fundamental human rights and wellbeing of citizens in the years to come. — Enrique Zapata and Constanza Gomez Mont

(Picture Credit: Unsplash)