AI Governance Retrospective Briefing — December 20, 2021
European and U.S. regulators laid out the AI guardrails Zeph Tech still follows, from the EU’s 2020 white paper through NIST’s 2021 AI risk management consultations.
Global regulators spent 2020 and 2021 sketching the guardrails Zeph Tech still references when advising on enterprise AI. Their white papers, executive mandates, and funding frameworks mapped accountability requirements before generative AI entered production roadmaps.
- 19 Feb 2020 — European Commission White Paper on Artificial Intelligence. Brussels’ risk-based blueprint coupled with the European data strategy defined “high-risk” AI obligations and foreshadowed conformity assessments across safety-critical deployments.
- 03 Dec 2020 — U.S. Executive Order 13960 on trustworthy AI. The order required federal agencies to inventory AI use, publish governance plans, and adopt the NIST AI Risk Management Playbook, signalling procurement expectations for explainability and privacy.
- 01 Jan 2021 — National AI Initiative Act of 2020. Congress codified the National AI Initiative Office, an interagency coordination structure that links NIST, NSF, DOE, and other regulators on standards, education, and evaluation infrastructure.
- 21 Apr 2021 — European Commission Artificial Intelligence Act proposal. The draft regulation converted the 2020 white paper into binding obligations, layering CE-mark style conformity assessments, incident reporting, and obligations for high-risk systems and providers of general-purpose AI.
- 29 Jul 2021 — NIST request for information on the AI Risk Management Framework. NIST opened a formal docket for industry input on managing AI risks, setting the foundation for the RMF workshops, profiles, and cross-sector evaluation tools used today.
- 03 Dec 2021 — OSTP and NSF request for information on a National AI Research Resource. The RFI scoped shared compute, data, and testing facilities so regulators could align public-private investments with responsible AI benchmarks.
For Zeph Tech’s AI pillar, these milestones still anchor model governance templates, supplier questionnaires, and product review cadences. Every new release is benchmarked against the compliance concepts first framed in 2020–2021.