Trusted rails for AI commerce
Venture capital investment in artificial intelligence-focused start-ups is accelerating to meet high demand.
South Africa led the continent with $610 million in investment during 2023, followed closely by Nigeria with $218-million and Kenya with $15 million.
This capital creates a fertile ground for innovation in multiple sectors. Realising this vast potential requires building secure, trusted digital rails for everyday commerce and data exchange.
The research highlights a significant opportunity to expand financial inclusion.
There are over 400 million unbanked or underbanked adults in sub-Saharan Africa.
Artificial intelligence (AI) can help bridge this divide by enabling new ways to assess risk and extend services.
Digital agents can assist consumers in securely sharing their permissioned data, such as mobile money usage history, with local financial institutions.
Users retain complete oversight of their personal information at all times.
These permissioned data exchanges sit alongside Mastercard’s core role as a trusted transaction network – linking identity, intent, authentication and settlement into a single, secure flow.
This consumer-led approach ensures data rights are respected while safely expanding access to vital financial services.
Businesses and consumers are looking for ways to reduce friction in their daily transactions.
Imagine a scenario where a small business owner in Johannesburg directs a digital agent to execute a pre-approved procurement within clearly defined parameters, such as preferred suppliers, price thresholds and delivery requirements.
The critical question is how the business owner can trust that agent to execute the payment securely and accurately.
Agentic commerce framework
Mastercard is addressing this exact challenge through specific frameworks designed for trusted, agent-initiated transactions.
This framework is supported by massive investments in network security.
Mastercard has invested approximately $11 billion in cybersecurity innovation since 2018.
The company already utilises generative AI to protect the broader financial ecosystem.
Its Decision Intelligence Pro tool analyses over one trillion data points to deliver a 200 per cent higher fraud detection rate.
In the United Kingdom, a similar solution detects 20 per cent more authorised push payment fraud, saving a potential £100 million annually.
Mastercard is now applying this deep expertise to establish the guardrails for a new era of agentic commerce.
The agentic commerce framework rests on strict pillars of consumer control and interface standards.
Any digital agent operating on the network must be fully verified and registered. The system relies on precise intent capture.
The human user always sets the initial parameters and financial boundaries.
If a transaction falls outside a predefined limit, the system pauses. If the user’s intent changes, the network requires immediate re-authentication.
Agents cannot authorise payments without explicit human consent and clearly defined validity windows.
This eliminates the risk of unapproved spending and keeps the consumer firmly in control.
To protect financial data during these interactions, the network deploys agentic tokenisation.
These controls ensure the user’s actual account details remain entirely hidden when the digital agent interacts with merchants or third-party services.
The framework guarantees total transparency alongside this privacy. Users maintain a clear, accessible record of exactly which agent acted on their behalf at any given time.
Responsible AI
These mechanisms are built entirely on foundational principles for responsible AI development. Mastercard prioritises trust, transparency, accountability and security in every product.
The company fosters this culture internally through its AI and Data Science Guild, which boasts approximately 2,000 members.
A broader AI Enthusiast Collective engages nearly 7,000 members across the organisation.
This deep internal expertise ensures that every new payment solution is rigorously tested and secured before reaching the market.
Stakeholders need a coordinated approach to navigate the future of digital finance. Fragmented regulatory landscapes and data silos remain significant challenges across the continent.
By building secure ecosystems for user-directed agents, the financial sector can work constructively within diverse regulatory environments to build interoperable, compliant ecosystems.
This careful, structured approach will safely unlock the true value of artificial intelligence for African commerce.
Loading ...