March 31, 2026 | By Deniz A. Johnson, COO/CIO at Stratyfy
It’s no surprise that global AI innovation is outpacing governance, leaving institutions building the plane while flying. And it’s not just a problem for banks; our lack of a unified regulatory infrastructure is the biggest risk to scaling AI safely.
Singular global AI regulation may be a pipe dream, but we can create a unifying framework for governance, with clear boundaries that allow industries to embrace AI with confidence. I believe that framework should be rooted in Responsible AI.
The Global AI Race: A Fragmented Landscape
Currently, the regulatory landscape is a complex patchwork. Over 72 countries have initiated AI policies, but they are far from uniform:
- Singapore was one of the first countries putting boundaries on AI and establishing best practices. At the World Economic Forum this past January, Singapore unveiled the Model AI Governance Framework for Agentic AI (MGF): the world’s first dedicated governance model for agentic AI systems.
- The EU also leads the charge with a strict, risk-based AI Act, positioning itself as promoter of a “rights‑driven” model to AI governance. The EU AI Act is the world’s most comprehensive AI regulatory framework, with a strict risk-based classification system and strong emphasis on transparency and prohibitions.
- China uses a highly centralized national model where local provinces follow strict Beijing-directed rules and five-year plans. It emphasizes institutional responsibility (e.g., ethics committees in enterprises and universities). With its top-down strategy, China has been able to quickly build out AI research parks, national computing infrastructure, and industry-specific AI deployment programs, all designed to meet the government’s goals.
- The United States continues to juggle a fragmented mix of NIST frameworks and state-level orders, with California, Colorado, and New York leading the way in terms of AI safety and anti-discrimination laws. With a National Policy Framework released this March, there has been significant executive pressure for a national lighter-touch approach and an AI Litigation Task Force created to challenge state laws on the grounds that they interfere with interstate commerce.
The Challenges of Global Innovation are Not New
AI adoption in today’s regulatory environment feels a lot like the GDPR scramble of early 2018, but with higher stakes and a much more complex technical layer. This regulation successfully collapsed a fragmented landscape of 28 different national laws into a single, cohesive framework.
At the time, each institution took its own approach to GDPR compliance. I was one of the early teams that went through GDPR, implementing a product in the EU and US simultaneously during my time at a large global financial institution in 2018.
Initially, many US companies initially thought they could ignore GDPR or bifurcate their services. Instead, we followed this unified framework and avoided a maintenance and support nightmare while rolling out the product faster. It was more cost-effective and efficient to make the product GDPR-compliant, rather than have distinct US and EU versions.
The lesson? It is infinitely more efficient to build toward one unified standard than to manage a dozen mediocre versions. Fragmentation doesn’t just slow down business; it’s an expensive way to dilute our impact.
The Emerging Framework: Responsible AI
As we saw with GDPR, the goal of these regulations shouldn’t be to slow us down, but to provide the frameworks that allow companies to be innovative (and profitable) while preserving user trust.
When the speed of AI innovation and adoption is exponentially faster than ever before…when it’s as easy as 6000 questions to bypass standard checks and manipulate an agent…when chatbot interactions are linked to teen suicides…it is clear that there is a need for a Responsible AI framework to help guide global safety, policy and innovation.
We are already seeing a global shift toward this, with countries worldwide integrating Responsible AI considerations into their frameworks:
- The OECD AI Principles established international consensus on AI as early as 2019, emphasizing inclusive growth, sustainable development, human-centered values, transparency, and accountability. These principles were updated in 2023/2024 and have been adopted by over 40 countries, most recently influencing landmark efforts like the EU AI Act.
- UNESCO released its Recommendation on the Ethics of Artificial Intelligence in 2021, which emphasized ethical governance and stewardship, robust data governance and protection, and comprehensive AI impact assessments to ensure human rights are protected.
- The NIST AI Risk Management Framework was introduced in 2023 to help organizations manage risks in AI systems. By defining “trustworthy AI” and providing guidance on how to address risks, the framework is designed to be flexible and apply to any organization, regardless of size or sector.
As AI innovation changes by the minute, governance rooted in Responsible AI should be a compass, not a tether.
Looking Ahead: Innovation Through Principles
The future of AI regulation must move away from vague policies and toward tangible guidelines and defined metrics, with Responsible AI as the unifying principle. This will allow industries to embrace AI in their own unique ways, leading to higher productivity and economic growth, while we focus on re-skilling our workforce to optimize the human-AI relationship.
What would that look like? It’s already being defined. The DNA of Responsible AI, and the basis for regulation such as the EU AI Act, the OECD AI Principles, and the NIST AI Risk Management Framework, is built on seven core pillars that were recently reaffirmed at the AI Impact Summit in Delhi in February 2026:
- Transparency & Explainability: Knowing why a system makes a decision.
- Fairness & Non-Discrimination: Ensuring AI doesn’t perpetuate bias.
- Accountability & Human Oversight: Keeping humans in the loop.
- Robustness, Security, & Safety: Protecting against manipulation and system failure.
- Privacy & Data Governance: Building on the foundations of GDPR.
- Risk-Based Classification: Treating high-stakes applications with appropriate scrutiny.
- Auditability & Continuous Monitoring: Ensuring systems remain safe over time.
Any technology is nothing without adoption. Unified regulation with Responsible AI at its core will help us build a sustainable, innovative AI economy that people actually want to live in.
Deniz Johnson is the COO/CIO at Stratyfy. This topic was originally presented at the Association for Financial Technology Spring Summit. If you’d like to discuss the future of global AI regulation and responsible frameworks, feel free to reach out at deniz@stratyfy.com.