The intersection of AI agents, blockchain, and everyday finance is no longer a distant vision—it's happening now, and fast. From agentic commerce that hides complex rails behind simple conversations, to trust stacks built on permission, proof, and verifiable receipts, the industry is shifting toward AI as the true interface for financial decisions. Users want outcomes in plain language ("pay this," "save that," "grow my assets safely"), not chains or wallets to manage.
Fresh off insights from Consensus Hong Kong 2026, where agentic payments and multi-rail routing dominated discussions, Olayimika Oyebanji recently sat down with Aurum Foundation CEO Bryan Benson to discuss the trust primitives needed for mass adoption, and why AI-native finance could finally pass the "grandma test" for billions. Let's dive in.
Can you briefly tell us about yourself and your route to AI and blockchain?
I’ve spent more than 27 years building and scaling fintech and digital asset businesses across multiple regions. My time as Managing Director at Binance made me realize that retail participation grew fast while institutional-grade tooling, risk controls, and liquidity intelligence stayed concentrated at the top end of the market. That’s what got me into blockchain infrastructure, DeFi rails, and the systems that can move money more efficiently.
AI became the next logical step because it keeps up with markets that run 24/7. Machine-scale monitoring and rules-based execution reduce emotional errors, respond faster to volatility, and make sophisticated trading behavior more usable for everyday investors. We are building this exact infrastructure to give individuals the same advantages professional desks rely on.
At Consensus Hong Kong 2026, agentic payments and AI agents were a dominant theme—how do you see agents evolving from payment tools to full AI finance interfaces?
That evolution is already underway. Agents are becoming the front end, while the underlying rails, wallets, chains, cards, bank transfers, and stablecoins, fade into the background. The architecture that makes this workable is permission and proof: clear mandates for what an agent can do, plus verifiable receipts for what actually happened.
That came through clearly in the Consensus Hong Kong conversations, including my discussion with Yat Siu and Tobias Bauer. The real shift is intent-based finance. Everyday users should be able to express a goal in plain language, then let software execute safely within defined boundaries. That is the “grandma test,” and it is a crucial step toward full AI finance interfaces.
Agents need bounded permissions, time limits, allowlists, confirmations, and a kill switch, then they need to show a readable trail of what was requested, what was executed, and where funds moved. We are designing around that model, AI-driven execution with clear constraints and visible receipts.
Onboarding friction (KYC) and chargeback risks remain major barriers to mainstream adoption—how can progressive onboarding and hybrid fiat/stablecoin rails help bridge traditional finance and DeFi?
Long KYC funnels drive significant drop-off. Research shows 68% of consumers abandon digital banking applications mid-onboarding due to complex processes. Progressive onboarding helps by matching verification depth to risk.
Platforms can start users with low limits and basic functionality, then unlock higher balances, transfers, and advanced features as verification increases. That gives users a faster first step while still keeping compliance and monitoring intact, especially in emerging markets..
Hybrid fiat and stablecoin rails address chargebacks and settlement uncertainty. Card payments are pull-based and reversible, which raises fraud and dispute exposure. Stablecoin settlement behaves more like a push payment, which can reduce chargeback pressure when it is paired with strong authentication, clear receipts, and sensible cooling-off rules for withdrawals.
In recent talks, you’ve discussed AI outperforming humans in trading and finance—what specific AI innovations are already demonstrating this edge?
The clearest edge comes from three innovations working together. The real edge is autonomous execution — systems that act like a disciplined desk. They watch liquidity, volatility, and spreads across venues, then size and route orders inside preset risk limits.
Then there's reinforcement learning, which has matured beyond “learn from the past” using post-training methods such as self-play simulations and verifiable reward checks, strategies learn decision rules that hold up when conditions change. Finally, models are now fusing more context, combining market data with signals like news sentiment, which helps them react earlier to regime shifts.
These tools have already beaten human-managed benchmarks. Following the deployment of AI-powered automation technology, private trading firms like BlueCrest gained 73% in 2025, significantly outpacing the 17.9% total return for the S&P 500.
With no single chain likely to dominate, how should platforms optimize for this in agentic commerce, and what mitigants address finality mismatches or fragmentation?
Platforms should optimize for intent and policy, then route execution across whatever rails fit the job best. In practice, that means the user gives one instruction, and the system handles chain selection, settlement path, and liquidity routing in the background across fiat rails, stablecoins, and multiple chains.
The key is risk-aware orchestration. Finality is different across networks, so agents need confirmation thresholds by chain and asset, exposure caps while transactions are pending, and fallback routes when liquidity or latency degrades. High-risk flows may also need escrow-style holds or delayed release rules.
A standardized receipt layer is just as important, because users and platforms need a clear audit trail that shows intent, route, execution, and settlement across fragmented infrastructure.
Your background scaling Binance in emerging markets brings deep experience in fiat on/off-ramps and inclusion—how are you applying those lessons for a seamless global access to AI-powered financial products?
We focus on building a complete financial stack that works the way people already manage money: mobile-first, cross-border, and across both fiat and crypto rails. That means combining payments, cards, wallets, and trading infrastructure with AI-driven execution tools, so users do not need separate apps and workflows for every task.
The lesson from emerging markets is that access alone is never enough. Products need local compatibility, clear risk controls, and a user experience that stays simple under volatile conditions.
We apply that through hybrid rails, progressive onboarding, and automation, which helps users manage exposure and act with more discipline. The goal is a system that feels familiar on the surface and institutional-grade in how it handles execution underneath.