- The AI Check In
- Posts
- 🛟 Why More US Banks Are Walking Away from MiFID II and EU AI Rules
🛟 Why More US Banks Are Walking Away from MiFID II and EU AI Rules
U.S. banks are backing away from EU-style AI governance as domestic regulators push outcome-based oversight. Leading firms are adopting modular, U.S.-first frameworks that speed up model approvals, reduce compliance drag, and sharpen audit readiness—replacing MiFID II controls with governance that fits their actual risk footprint.
👋 Welcome, thank you for joining us here :)
The AI Check In is your weekly power play to navigate AI governance and strategy in banking and finance.
What to expect this edition:
🛟 Need to Know: The EU is Fading. U.S. AI Governance is Splintering.
🥷 Deep Dive: Why U.S. Banks Are Ditching EU Compliance Frameworks
⚔️ Instruments of Mastery: IBM OpenPages AI Governance
📈 Trends to Watch: From Harmonization to Fragmentation: AI Regulation Goes Local
Rewrite the rules. Outmaneuver competitors. Stay two steps ahead of regulators.
Let the games begin.
🛟 Need to Know: The EU is Fading. U.S. AI Governance is Splintering.
The AI compliance game has changed. U.S. financial institutions are rapidly shifting away from EU-style governance templates as regulators intensify enforcement around domestic obligations. While MiFID II once set the tone for global standards, its operational relevance is shrinking—particularly for banks without direct exposure to European markets.
Introduced in 2018, the EU’s Markets in Financial Instruments Directive II (MiFID II) significantly reshaped transparency, research funding, and risk governance in global financial markets. Many U.S. banks voluntarily aligned to protect European market access. Today, that default posture is eroding.
U.S. regulators are stepping into the void. The OCC, FDIC, and CFPB now prioritize outcome-based AI oversight—requiring fairness, traceability, and explainability—without prescribing rigid technical mandates. Simultaneously, state-level laws such as Colorado’s CAIA (effective 2026) and California’s proposed AI bills are creating a fragmented compliance landscape that renders one-size-fits-all frameworks obsolete.

For boards preparing 2025 audit cycles, the message is clear: outdated global alignment is no longer sufficient. Institutions must demonstrate jurisdiction-specific AI governance or risk being caught short under scrutiny.
Emerging Best Practices:
Dynamic jurisdiction mapping: Governance protocols are now mapped by legal exposure, not harmonized globally by default.
Principles-first frameworks: AI oversight focuses on fairness, explainability, and resilience across core U.S. mandates.
Audit integration tooling: Platforms like IBM OpenPages and ServiceNow are being used to document and monitor AI risk posture without defaulting to EU-style compliance overhead.
🥊 Your Move:
Reassess EU alignment thresholds: Only maintain EU-style AI governance controls if models are deployed in or serve EU markets. Otherwise, modular, local-first compliance may be more efficient.
Map AI compliance by jurisdiction: Maintain a live inventory of where each AI system operates and which laws apply—particularly as Colorado, California, and Illinois develop divergent AI rules.
Prioritize audit-readiness over prescriptive rules: Focus on traceability, fairness documentation, and explainability logs that satisfy U.S. audit and enforcement expectations, without mirroring EU technical mandates.
🥷 Deep Dive: MiFID II vs. U.S. Deregulation — Where Should AI Compliance Efforts Be Prioritized?
Why U.S. Banks Are Ditching EU Compliance Frameworks
MiFID II was once the global benchmark for AI and algorithmic governance. Today, it’s a cost center that many U.S. institutions are leaving behind. As regulatory enforcement aligns more closely with jurisdictional boundaries, strategic banks are abandoning blanket compliance in favor of modular frameworks that deliver speed, defensibility, and capital efficiency.

MiFID II’s Shrinking Relevance
Introduced in 2018, MiFID II reshaped algorithmic risk controls, trading transparency, and research funding norms across Europe. For years, U.S. banks adopted voluntary alignment to preserve EU market access or avoid reputational exposure.
That era is over:
U.S. banks without regulated EU operations are stepping back from MiFID II-aligned controls.
There is minimal recent enforcement of MiFID II against U.S. institutions with limited European exposure.
Internal compliance teams are reallocating resources toward frameworks that reflect U.S. regulatory expectations.
This recalibration reflects a simple truth: global harmonization is outpaced by regional divergence. Maintaining unnecessary EU-style structures no longer passes the capital allocation test.
The U.S. Model: Risk-Driven, Audit-Ready
Domestic regulators are advancing a more flexible, accountability-focused approach:
OCC & FDIC: Emphasize traceable model documentation and ongoing validation under broader model risk frameworks.
CFPB Circular 2022-03: Requires clear, specific factors in adverse action notices—even for black-box models.
State Legislation: Colorado’s CAIA mandates AI impact assessments and public disclosures; similar bills are advancing in California and Illinois.
These frameworks demand fewer global abstractions—and more localized control. The institutions adapting fastest are streamlining compliance, speeding up approvals, and reducing regulatory capital exposure linked to operational risk ratings.