AI agents pose new risks to Nigerian banks, CBN warns – Businessday NG

Artificial intelligence (AI) agents are creating new cybersecurity risks for Nigerian banks, with the Central Bank of Nigeria (CBN) warning financial institutions against granting automated systems excessive access, authority and autonomy over critical operations.

Advertisement
Advertisement

Rakiya Yusuf, director of the Payments System Supervision Department at the CBN, said AI applications connected to customer information, institutional infrastructure and data systems could become entry points for cyberattacks if appropriate safeguards and access controls were not established.

She warned that AI agents allowed to perform functions without adequate boundaries could amplify the consequences of compromised credentials, malicious prompts and human errors.

“For safety and resilience, AI should never be given more access, authority, or autonomy than the implementing institutions can safely manage,” Yusuf said at the 2026 Annual Cybersecurity Conference of the Information Security Society of Africa-Nigeria (ISSAN).

She said AI was already helping financial institutions strengthen transaction monitoring, identify unusual customer behaviour, detect fraud patterns, improve customer authentication and support anti-money laundering controls.

The technology was also improving the ability of financial regulators to analyse large volumes of transaction data through supervisory technology tools.

Advertisement

However, Yusuf said the same capabilities strengthening cybersecurity defences were also empowering criminals to conduct attacks faster and on a much larger scale.

“AI risk does not only come from criminals. There is also the risk from the AI implementation method that is adopted by legitimate organisations,” she said.

Advertisement

According to her, financial institutions must recognise that cybersecurity threats are no longer limited to external criminal activities but can also emerge from poorly governed technology deployments within their own operations.

She urged banks to adopt security-by-design principles, strengthen identity and access controls, improve intelligence sharing, test their resilience and maintain human oversight of AI systems.

Advertisement
Advertisement

Yusuf also called on financial institutions to secure their AI supply chains, protect institutional data and ensure that employees and management understand the technology before deploying it.

Blaise Ijebor, director of the Risk Management Department at the CBN, said artificial intelligence was lowering the barriers to entry into cybercrime by enabling individuals with limited technical expertise to execute sophisticated attacks.

He said AI could generate convincing corporate emails, imitate people’s writing styles and voices, create synthetic identities, write malicious code and automate several stages of cyberattacks.

“AI may be the greatest productivity tool ever given to cybercriminals,” Ijebor said, explaining that the technology was providing attackers with the same advantages of speed, scale and sophistication that businesses were seeking to achieve through digital transformation.

Citing an investigation by AI company Anthropic, Ijebor said AI had reportedly executed between 80 and 90 percent of a large-scale cyberattack targeting approximately 30 organisations, including technology companies, financial institutions and government agencies.

He said the development marked a shift from criminals using AI to assist cyberattacks to human-directed AI agents independently executing most stages of an attack.

The CBN official warned that the implications were particularly significant for Nigeria’s rapidly expanding digital financial ecosystem, where instant payments could increase the speed at which stolen funds are transferred.

“Instant payment can also mean instant losses,” he said.

Ijebor explained that criminals could combine AI-generated deception, information gathered from social media, stolen credentials and weaknesses in digital infrastructure to manipulate employees into authorising fraudulent transactions.

He illustrated the threat with a scenario in which an employee receives a WhatsApp message and telephone call apparently from a company’s chief executive, instructing an urgent transfer of N50 million.

With AI capable of cloning voices and generating convincing videos, he said traditional verification methods based on recognising a person’s voice, appearance or writing style were becoming increasingly unreliable.

“The critical question is no longer simply, does this look genuine? The question now is, can I independently prove that this is genuine?” he said.

He therefore urged banks to strengthen identity verification procedures, insisting that urgency, confidentiality and instructions not to call back should never override established authentication controls.

Beyond preventing attacks, Ijebor said financial institutions must build the capacity to contain breaches, maintain critical operations and recover without losing customer confidence.

“Prevention matters, but prevention alone is not enough,” he said, adding that banks must regularly test backup systems, conduct cyberattack simulations and assess vulnerabilities across their technology suppliers.

He also cautioned against poorly governed AI applications, including employees uploading confidential institutional documents into publicly available AI platforms without understanding how the information could be accessed or reused.

David Isiavwe, president of the Information Security Society of Africa-Nigeria (ISSAN), said artificial intelligence was expanding the cybersecurity threat landscape by enabling criminals to execute attacks with greater speed and sophistication.

He warned that deceptive messages, cloned voices, synthetic videos and social engineering attacks were becoming increasingly difficult to identify, raising the risks for financial institutions and other organisations operating in interconnected digital environments.

According to him, cyberattacks could have consequences extending beyond the institutions or countries directly affected, making cybersecurity a shared responsibility across the digital ecosystem.

“Effective cyber defence depends on cooperation, collaboration and ecosystem thinking,” Isiavwe said, calling for stronger public-private partnerships, information sharing and prompt action on identified cyber threats.

He said trust remained central to the digital economy, stressing that customers must be able to rely on the security and integrity of the financial and other digital services they use.

Dele Alabi, president of the Chartered Institute of Bankers of Nigeria (CIBN), said AI was improving efficiency and innovation in financial services but was also equipping criminals with tools to create synthetic identities, deepfakes and automated social engineering attacks.

He warned that a single cyber breach in Nigeria’s highly interconnected financial ecosystem could spread across institutions and markets, creating risks beyond the organisation directly affected.

According to Alabi, Nigeria processed approximately 47.88 billion electronic payment transactions valued at N3.46 quadrillion in 2025, underscoring the growing importance of cybersecurity to the country’s financial system.

He disclosed that industry efforts helped reduce electronic payment fraud losses by more than 50 percent to N25.85 billion in 2025, from N52.26 billion in 2024.

However, he cautioned that the figures could underestimate the scale of the problem because some fraud incidents were not reported.

Alabi said cybersecurity had moved beyond being an information technology concern to becoming a strategic management issue and a critical pillar of financial stability.

He called on bank boards to invest in cyber resilience, management teams to integrate cyber risks into enterprise-wide decision-making and employees to adhere strictly to security protocols.

He also emphasised the importance of ethical leadership, employee vigilance and collaboration among financial institutions in strengthening the security of Nigeria’s digital economy.

The three speakers stressed that financial institutions must balance technological innovation with security, accountability and resilience as AI adoption accelerates.

They maintained that while preventing every cyberattack may be impossible, banks must ensure that breaches do not disrupt critical financial services or undermine public confidence in digital payments.

Hope Moses-Ashike

Hope Moses-Ashike is an Associate Editor, Banking and Finance, with more than a decade of experience reporting on Nigeria’s financial system and broader economy. She closely tracks market movements, monetary policy decisions, company disclosures, regulatory actions, economic indicators, and global developments, and interprets what they mean for businesses, investors, policymakers, and households. Her reporting helps readers understand complex issues such as inflation trends, foreign exchange market dynamics, interest rate decisions, bank performance, and investment risks.

She also covers major international events and periodically travels to Washington, D.C., to report on the World Bank/IMF Spring and Annual Meetings.
Her dedication to financial journalism has earned her multiple recognitions and invitations to high-level professional development programmes. She is an alumna of the International Visitors Leadership Programme (IVLP) in the United States and holds an Advanced Financial Journalism Certificate from the Press Association Training in London, UK. Her other notable achievements include completing the Lagos Business School CMC Programme, the Bloomberg Media Africa Initiative Programme, and a Master Class in Journalism at Rhodes University in South Africa.

Advertisement

Source:

Businessday NG