The rapid development of Artificial Intelligence (AI) is creating a widening gap between what AI systems can do and the rules, institutions and safeguards designed to control them, prompting some of the technology industry’s most influential figures to call for stronger intervention.
Bill Gates, the Microsoft co-founder who has spent years promoting AI’s potential to transform healthcare, education and productivity, has become increasingly concerned about the speed of its development.
In a 2026 essay and interview, Gates said he was worried about AI-enabled hacking, biological weapons, fraud, job losses and its effects on children, while arguing that the industry could not regulate itself effectively.
Gates’ concerns come as AI companies continue to push the boundaries of what their systems can accomplish. OpenAI has this month called for international technical standards for frontier AI, including systems capable of recursive self-improvement, arguing that AI development needs to be paced so that safety research can keep up with capability growth.
The tension is becoming increasingly difficult to ignore, in that, the same companies and technology leaders warning about AI risks are also competing aggressively to build more capable systems.
Read also: FG, UNHCR target $10bn private investment to rebuild displacement-hit regions
From promise to a governance problem
For years, AI safety was largely treated as a technical problem; whether engineers could prevent models from producing harmful content or behaving unpredictably.
The challenge is now much broader. More capable AI systems are being integrated into coding, cybersecurity, research, customer service, education and decision-making. That means failures can increasingly move beyond a chatbot producing a wrong answer to systems taking actions, processing sensitive information or influencing consequential decisions.
Gates has pointed to the emergence of AI systems capable of completing increasingly complex workplace tasks and participating in sophisticated cyber activity as evidence that the technology is moving faster than he expected. He has also argued that governments need to play a stronger role because the industry cannot be relied upon to police itself.
OpenAI CEO Sam Altman has similarly acknowledged the scale of the risk. Altman said even a 10 percent chance of AI contributing to human extinction would be unacceptable, while calling for greater attention to safety and alignment.
The concerns are not limited to existential scenarios. Deepfakes, automated fraud, cyberattacks, misinformation, privacy violations and potential job displacement represent more immediate problems for governments and businesses.
For Nigeria, the question is particularly important because the country is trying to accelerate AI adoption while simultaneously building the regulatory capacity needed to manage it.
Nigeria is building its own guardrails
Nigeria has moved beyond discussions about whether it should participate in the AI economy.
The Federal Government’s National Artificial Intelligence Strategy, finalised in 2025, identifies the need for a national AI policy framework and a national AI risk management framework to identify, assess and mitigate safety and security risks associated with AI systems.
The strategy also recognises risks involving bias, privacy, transparency, job automation and the deployment of AI in high-stakes sectors such as healthcare, finance and security.
Kashifu Inuwa Abdullahi, director-general of the National Information Technology Development Agency (NITDA), has argued that Nigeria’s approach must remain flexible because AI is changing too quickly for static regulation.
NITDA has described its approach as a “Regulatory Intelligence Framework”, built around awareness, intelligence and dynamism. Inuwa has said regulators can combine conventional rules with a more flexible approach in which developers build use cases and regulators establish guardrails and best practices around them.
That approach reflects a central challenge facing regulators globally: writing rules for technologies whose capabilities may change faster than the legislative process.
NITDA is also pushing Nigeria towards greater technological sovereignty. In September, Inuwa warned against excessive reliance on AI systems developed and controlled outside Nigeria, arguing that dependence could undermine the country’s digital and economic sovereignty as AI becomes embedded in critical sectors.
The agency is simultaneously developing local capability. Its National Centre for Artificial Intelligence and Robotics is supporting N-ATLAS, an open-source multilingual large language model designed around Yoruba, Hausa, Igbo and Nigerian-accented English. A National AI Innovation Challenge launched this month is intended to move N-ATLAS from model availability towards practical national infrastructure and applications.
Trust becomes the missing layer
For Nigerian policymakers, however, building AI capability is only one side of the equation.
Bosun Tijani, minister of communications, innovation and digital economy, has been advancing the idea of a National AI Trust as Nigeria seeks to formalise oversight around AI adoption and responsible innovation.
At a National AI Trust convening in London in June, Tijani brought together government officials, AI experts, businesses, academics and public-interest groups to discuss how the proposed trust could guide the next stage of Nigeria’s AI adoption and investment.
He was also appointed a founding member of the AI for Good Global Commission, an international initiative involving governments, technology leaders and international organisations focused on trusted and inclusive approaches to AI.
The Nigerian approach increasingly recognises that AI governance cannot sit within one regulator.
In July, the Federal Ministry of Communications, Innovation and Digital Economy directed the NCC, NITDA and the Nigeria Data Protection Commission to work towards a harmonised national policy and governance framework as the boundaries between telecommunications, AI, online platforms, cybersecurity and data protection become increasingly blurred.
Data protection becomes an AI issue
For Vincent Olatunji, chief executive of the Nigeria Data Protection Commission, the AI debate is inseparable from the question of who controls the data used to train and operate these systems.
In February, the NDPC joined international data protection authorities in warning about AI-generated images and videos involving identifiable people without their knowledge or consent.
The commission said AI developers must build safeguards into their systems, provide transparency about capabilities and establish mechanisms for quickly removing harmful personal content.
Olatunji also said Nigeria may need to review its data protection legislation to address the rapid development of AI, robotics and big data, arguing that the digital environment has evolved significantly since the Nigeria Data Protection Act was enacted in 2023.
This points to another weakness in the global AI race: even where laws already exist, they may not have been written with increasingly autonomous AI systems in mind.
Industry faces its own accountability test
Technology companies are also under pressure to demonstrate that responsible AI is more than a policy statement.
Microsoft West Africa’s director of government affairs, Nonye Ujam, said Nigeria had made progress through its National AI Strategy and regulatory reforms but needed to move from policy development to implementation.
She identified governance frameworks, infrastructure and institutional capacity as critical to scaling AI while maintaining fairness, security, transparency and accountability.
Ujam has separately argued that Nigeria’s public-sector workforce will be critical to determining whether AI delivers economic benefits, stressing that government institutions need the skills to understand, deploy and govern the technology.
That institutional question may become as important as the technology itself. A regulator cannot effectively supervise an AI system it does not understand. A government cannot meaningfully audit automated decisions without technical capacity. And businesses cannot protect customers from AI-driven risks if their employees lack the skills to identify them.
Read also: 127 Law students participate in NDDC moot court, mock trial competition
The race versus the guardrails
The emerging global debate is therefore no longer simply about whether AI should be developed. It is about whether the institutions responsible for managing its consequences can move at the same speed as the companies developing it.
The commercial incentives are enormous. Companies have strong reasons to keep improving models, while countries fear losing technological and strategic ground if they slow down. That creates a difficult incentive structure: every company has an individual reason to move quickly even when the collective risks of moving too quickly are difficult to measure.
OpenAI’s recent call for international technical standards reflects this dilemma. The company argued that coordinated standards and incident reporting are needed to prevent fragmented approaches to frontier AI.
For Nigeria, the stakes are different but connected. The country needs AI to improve productivity, create new businesses and strengthen public services. But it also needs systems that protect personal data, preserve human oversight, address local languages and biases, and ensure that AI-generated decisions can be challenged.
The central lesson from the warnings by Gates and other technology leaders is therefore not that AI development should stop. It is that capability alone cannot determine the direction of the technology.
Nigeria’s challenge will be to build its AI economy while ensuring that regulation, skills, data protection, local infrastructure and accountability develop alongside it.
As AI moves from a tool people consult to systems capable of performing increasingly complex tasks, the definition of responsible innovation is also changing. The race is no longer only to build more powerful AI. It is also to build the institutions capable of keeping up with it.
