When "AI for Good" Goes Wrong

This guest article is adapted from the author's new book From Pessimism to Promise: Lessons from the Global South on Designing Inclusive Tech, published by MIT Press.

What do AI-enabled rhino collars in South Africa, computer-vision pest-detection drones in the Punjab farmlands, and wearable health devices in rural Malawi have in common?

These initiatives are all part of the AI for Good movement, which aligns AI technologies with the United Nations sustainable development goals to find solutions for global challenges like poverty, health, education, and environmental sustainability.

MIT Press

The hunger for AI-based solutions is understandable. In 2023, 499 rhinos were killed by poachers in South Africa, an increase of more than 10 percent from 2022. Several farmers in Punjab lost about 90 percent of their cotton yield to the pink bollworm; if the pest had been detected in time, they could have saved their crops. As for healthcare, despite decades of effort to boost the numbers of healthcare practitioners in rural areas, they continue to migrate to cities.

What makes AI good," though? Why do we need to preface AI applications in the Global South with morality and charity? And will noble intent translate to making AI tools work for the majority of the world?

A Changed Reality

The fact is, the Global South of decades ago does not exist.

Today the countries in the Global South are more confident, more entrepreneurial, and are taking leadership to pioneer locally appropriate AI tools that work for their people. Startups understand that the success of new tech is contingent on leveraging local knowledge for meaningful adoption and scaling.

The old formula of innovate in the West and disseminate to the rest" is out of sync with this new reality. While the West holds onto its old missionary zeal, the South-South collaboration continues to grow, sharing new tech and building AI governance. What's more, some tech altruism initiatives have come under scrutiny as they obfuscate their data extraction activities, making them more transactional than charitable.

The Market for Tech Altruism

In August, the European Union's legal framework on AI, the AI Act, entered into force. Its measures are meant to help citizens and stakeholders optimize these tools while mitigating the risks. There is little mention of AI for good in their documents; it's simply the default. Yet, as we shift from the Global North to the South, morality kicks in.

Tech altruism underlines this shift. Many of the AI for Good initiatives are funded by tech philanthropists in partnership with global aid agencies. Doing good manifests in piloting tech solutions, with the Global South as a live laboratory. A running joke with development workers is that their field suffers from pilotitis," an acute syndrome of pilot projects that never scale. The Global South is typically viewed as a recipient, a market, a beneficiary for techno-solutionism.

Take AI collars for rhinos. The Conservation Collar initiative in South Africa, for example, detects abnormal behavior, and these signals are sent to an AI system that computes the probability of risk. If it determines that the animal is at urgent risk, the rangers can hypothetically act immediately to stop the poaching. But when my team investigated the ground realities, we found that rangers face numerous obstacles to fast action, including dirt roads, old vehicles, and long distances. Many rangers had not been paid in months, and their motivation was low. And to top all this, they faced an armed militia protecting a multibillion rhino trade business.

In Punjab, drones with computer vision can guide farmers to detect pests before they destroy crops. The Global Alliance for Climate-Smart Agriculture funds projects involving many such AI-enabled technologies as farmers face the vagaries of the climate crisis. However, detection is just one part of a larger problem. Farmers struggle with poor quality and unaffordable pesticides, loan sharks, the vulnerabilities brought on by monocropping, and water scarcity. Agricultural innovators complain that there are few early adopters of their tech, however good their tools may be. Afterall, young people in the Global South increasingly don't see their future in farming.

Meanwhile, we've seen philanthropies such as the Bill & Melinda Gates Foundation launch grand challenges for AI to help alleviate burdens on African healthcare systems. This has resulted in winners such as IntelSurv in Malawi, an intelligent disease surveillance data feedback system that computes data from smart wearables. Yet, even with hundreds of patents for such devices being registered every year, they're not yet capable of consistently capturing high-quality data. In places like Malawi, these devices may become the single source of training data for healthcare AI, amplifying errors in their healthcare system

The fact is, we can't really solve problems with AI without accompanying social reforms. Building proper roads or paying your rangers on time is not an innovation, it's common sense. Likewise, whether it's in the healthcare or the agricultural sector, people need social incentives to adopt these technologies. Otherwise these AI tools will remain in the wild, and won't be domesticated.

Data Is Currency

Tech altruism has increasingly become suspect as AI companies are now facing an acute data shortage. They're scrambling for data in the Global South, where majority of tech users live. Take, for instance, the case of Worldcoin, co-founded by OpenAI CEO Sam Altman. It plans to become the world's largest privacy-preserving human identity and financial network, giving ownership to everyone." Worldcoin started as a nonprofit in 2019 by collecting biometric data, mostly in Global South countries, through its orb" device and in exchange for cryptocurrency. Today, it's a for-profit entity and is under investigation by many countries for its dubious data-collection methods.

The German nonprofit Heinrich-Boll-Stiftung recently reported on the aggressive growth of digital agricultural platforms across Africa which promise farmers precision agriculture and increased yields via AI-enabled apps. Yet these apps often provide corporations with free access to data about seeds, soil, crops, fertilizers, and weather from the farms where they're used. Corporations can use AI analytics to weaponize this information, perhaps creating discriminatory agricultural insurance policies or micro-targeting ads for seeds and fertilizers. Similarly, in the health care sector, the Center for Digital Health at Brown University has reported on the selling of personal health data to third-party advertisers without user consent.

The problem is that, unlike private companies that are compelled to follow the law, altruistic initiatives often succeed in circumventing regulations due to their charitable" intent. Almost a decade ago, Facebook launched Free Basics, which provided access to limited internet services in the Global South by violating net neutrality principles. When India blocked Free Basics in 2015, Mark Zuckerberg appeared shocked and remarked, Who could possibly be against this?"

Today we ask, who could possibly get on board?

From Paternalism to Partnerships

As of 2024, according to one estimate, the Global South contributes 80 percent to global economic growth. Close to 90 percent of the world's young population reside in these regions. And it has become a vital space for innovation. In 2018, China entered the global innovation index rankings as one of the top twenty most innovative countries in the world. India's government has set up its tech stack," the largest open source, interoperable, and public digital infrastructure in the world. This stack is enabling entrepreneurs to build their products and services away from the Apple and Google duopoly that constrains competition and choice.

Despite the Global South demonstrating its innovative prowess, the imitator label remains sticky. This perception often translates to Western organizations treating Global South countries as beneficiaries, and not as partners and leaders in global innovation.

It is time we stop underestimating the Global South. Instead, Western organizations should channel their energies by looking at how different consumers can help to rethink opportunity, safeguards, and digital futures for the world's majority. Inclusion is not an altruistic act. It is an essential element to generating solutions for the wicked problems that humanity faces today.

In designing new tech, we need to shift away from morality-driven design with grandiose visions of doing good. Instead, we should strive for design that focuses on the relationships between people, contexts, and policies.

Designers, programmers, and funders can benefit from listening to what users and entrepreneurs in the Global South have to say about AI intervening in their lives. And policymakers should bury the term AI for Good."

Media outlets must stop debating whether tech alone can solve the world's problems. The real contextual intelligence we need won't come from AI, but from human beings.