EU AI Act brings compliance demands for private capital

The EU AI Act’s high-risk regime takes effect on 2 August 2026, bringing risk management, documentation, transparency and human oversight requirements for covered systems. Relevant uses in private capital include assessing the creditworthiness of natural persons and employment-related screening. Breaches of prohibited-use rules can attract fines of up to EUR 35 million or 7% of worldwide annual turnover, whichever is higher. Breaches of high-risk system obligations carry penalties of up to EUR 15 million or 3% of turnover.

Regulatory pressure extends beyond Europe. The UK, Singapore, Hong Kong, Australia and Gulf financial centres are developing or applying expectations around model governance, validation and oversight. The United States remains fragmented across securities regulation, state AI laws, sanctions and investment reviews. European fund managers also face overlapping portfolio reporting and sustainability disclosure requirements.

Operational priorities include maintaining an inventory of AI systems with risk classifications, named owners and validation records; documenting human review; assessing vendors; and strengthening investor-facing disclosures. A central requirement is decision provenance: an attributable record of the data considered, AI outputs used, human overrides and investment rationale. Capturing that evidence within investment workflows can support regulatory submissions and investor due diligence, while reducing reliance on retrospective reconstruction. Weak governance can also create fundraising delays, transaction obstacles and reputational exposure.