AI Adoption in Developing Economies: Key Strategies and Challenges

•

 11 min video

•

 4 min read

YouTube video ID: SSDLLeJiP00

Source: YouTube video by Stanford Graduate School of Business — Watch original video

PDF

The discussion around Artificial Intelligence (AI) differs significantly between advanced and developing economies. In advanced economies, the focus is on the "race to the frontier"—who will lead in AI development. For most developing countries, however, the primary concern is how to adopt existing AI technologies. This adoption must be balanced with adaptation, as AI learns from data that needs to reflect local languages, challenges, and conditions.

Government Actions in Developing Economies

Governments in developing countries can take three pragmatic steps:

  1. As Enablers of AI: Prioritize foundational investments in infrastructure and skills. This means focusing on essential services like electricity and internet connectivity rather than immediately investing in data centers.
  2. As Users of AI: Identify and implement AI applications in critical public services such as agricultural extension, education, and health. Governments should evaluate these solutions, scale up successful ones, and discontinue those that are not effective.
  3. As Regulators of AI: Address the risks associated with AI, including biases and malicious uses like cyber threats. Given resource constraints, governments should start with practical measures. This includes utilizing voluntary industry standards for AI and leveraging existing laws and regulations for cybersecurity and privacy, adapting them to address AI-specific challenges.

AI Adoption Trends in Developing Economies

Initial research into AI adoption in developing economies revealed some unexpected insights. While there is significant excitement about AI's potential to transform services like education and healthcare, particularly in regions with shortages of teachers and doctors, the current reality is different.

Much of the present AI activity in developing economies focuses on "back office" government functions. Governments are using AI for:

  • Predictive analytics
  • Decision support
  • Geospatial intelligence
  • Early warning systems
  • Smart agricultural practices

These applications often rely on more traditional, predictive AI models (older machine learning styles) rather than the generative AI models that receive more attention today.

Diffusion and Intensity of AI Use

Historically, newer technologies have reached developing countries faster than older ones. For example, steam power took 80 years to reach lower-income countries, electricity 40 years, and the internet 20 years. AI, particularly generative AI, continues this trend; middle-income countries accounted for about half of ChatGPT's worldwide traffic within six months of its launch.

However, while AI is reaching developing economies quickly, the intensity of its use remains a challenge. There's a growing gap in how deeply these technologies are integrated. For instance, approximately 25% of internet users in high-income countries use ChatGPT, compared to only about 5% in developing economies. This indicates that while the technology is accessible, its widespread and meaningful adoption is still limited.

AI is a general-purpose technology, similar to steam power, electricity, and the internet. Its diffusion is often slow because it requires complementary elements that are frequently missing. These include:

  • Infrastructure
  • Skills
  • Institutions
  • Organizational practices within businesses and governments

Significant changes are needed across these areas before AI can be fully integrated into how work is done.

Impact on Jobs

The conversation around AI's impact on jobs also differs between advanced and developing economies. In advanced economies, the concern is often about mass unemployment due to AI displacement. In developing economies, where a larger proportion of the workforce performs manual rather than cognitive tasks, the immediate disruption is less significant.

Less than 10% of jobs in developing economies are currently susceptible to automation by AI. Instead, AI is more likely to complement existing work. Developing countries often face shortages of professionals, such as doctors, agricultural experts, and skilled public sector officials. AI can help bridge these skill gaps, making these economies more productive.

While immediate job losses may be less severe, there could be future challenges related to structural transformation. Knowledge-intensive, middle-class jobs in developing countries might become more automated over time. However, as with past technological advancements, new tasks and occupations will emerge. Developing economies have some time to prepare for these changes by focusing on new skills and job creation.

Digital and Analog Complements

To accelerate development through AI, both digital and analog complements are crucial.

Digital Complements:

  • Local language data: Essential for training AI tools to be relevant to local contexts.
  • Internet connectivity: As many people as possible need access to the internet.
  • Digital devices: Access to devices that can harness AI's potential.

While these digital complements are undoubtedly vital, the importance of analog complements is often overlooked. Even as AI becomes more advanced, foundational elements remain critical. For example, a survey of businesses across seven developing economies revealed that the top reasons for not adopting or expanding AI solutions were:

  • Lack of knowledge, awareness, or information: Businesses were unaware of what AI could do for them.
  • Lack of buying power/access to finance: Inability to afford AI solutions.

These "analog" factors—foundational skills, information awareness, and access to finance—are as crucial as digital infrastructure. Governments must remember that while focusing on AI, data centers, and local language data, they should not neglect these fundamental building blocks for leveraging AI's potential.

  Takeaways

  • Advanced economies chase AI frontiers, while developing nations focus on adopting existing AI tools and tailoring them to local languages, challenges, and conditions.
  • Governments can act as enablers, users, and regulators by investing in basic infrastructure, deploying AI in public services, and applying voluntary standards and existing laws to mitigate bias and cyber risks.
  • Current AI use in developing countries centers on back‑office functions such as predictive analytics, decision support, geospatial intelligence, early warning systems, and smart agriculture, relying mostly on traditional machine‑learning models.
  • Although AI reaches developing markets quickly—evidenced by half of ChatGPT traffic coming from middle‑income nations—the depth of usage is low, with only about 5% of internet users in these economies regularly using ChatGPT.
  • Successful AI integration requires both digital complements (local language data, connectivity, devices) and analog complements (knowledge, awareness, financing), as lack of these fundamentals remains the top barrier for businesses.

Frequently Asked Questions

What are the three pragmatic steps governments in developing economies should take to support AI?

Governments should act as enablers, users, and regulators of AI. As enablers they invest in basic infrastructure and skills such as electricity and internet; as users they pilot AI in public services like agriculture, education, and health and scale successful pilots; as regulators they address bias and cyber threats by adopting voluntary standards and adapting existing privacy and cybersecurity laws.

Why is AI adoption intensity lower in developing economies despite rapid diffusion?

Adoption intensity remains low because many complementary elements are missing. Developing nations often lack reliable electricity, broadband connectivity, skilled personnel, and institutional frameworks, and businesses suffer from limited awareness and financing for AI projects. Without these digital and analog supports, AI tools are accessed but not deeply integrated into daily workflows.

Who is Stanford Graduate School of Business on YouTube?

Stanford Graduate School of Business is a YouTube channel that publishes videos on a range of topics. Browse more summaries from this channel below.

Does this page include the full transcript of the video?

Yes, the full transcript for this video is available on this page. Click 'Show transcript' in the sidebar to read it.

Helpful resources related to this video

If you want to practice or explore the concepts discussed in the video, these commonly used tools may help.

Links may be affiliate links. We only include resources that are genuinely relevant to the topic.

Full transcript is not shown on this page

This page focuses on the summary and original notes. For full verification, refer to the original YouTube video.

PDF