Other AI Regulatory Efforts

How current Washington State and U.S. efforts compare to a coherent, 13-pillar AI governance framework.

Short answer: yes, both Washington State and the federal government have launched AI-related efforts — task forces, executive orders, and targeted bills — but none of them yet add up to a single, unified “omnibus” AI law. They cover scattered pieces of the landscape your 13-pillar framework treats as a whole.

Below is a narrative summary first, then a more structured comparison to the 13 pillars.

Status Check

What’s Been Done So Far

Washington State

In 2024, the Washington legislature created a statewide AI Task Force to study how AI is being used, what risks it creates, and what kinds of rules or safeguards might be needed. The Attorney General’s office oversees this work.

The task force has a broad mandate. It is supposed to review how existing federal, state, and local laws apply to AI, identify high-risk applications, define key terms, propose guiding principles, recommend transparency and anti-bias protections, and suggest enforcement options and ongoing oversight mechanisms.

At this point, the group is still in the “work in progress” stage: holding meetings, gathering input, and preparing recommendations. There is not yet a standalone, finalized “Washington State AI code” that governs AI across the board.

Federal Level (United States)

At the federal level, the story is unfolding in stages. In 2023, Executive Order 14110 tried to set a nationwide direction for “safe, secure, and trustworthy” AI. It told federal agencies to work on transparency rules, safety tests, civil-rights protections, national security safeguards, and more.

In early 2025, that order was rescinded and replaced by a new one: Executive Order 14179, “Removing Barriers to American Leadership in Artificial Intelligence.” The new order emphasizes innovation and competitiveness and rolls back some of the earlier guardrails. Supporters see it as pro-growth; critics see it as weakening safety protections.

Meanwhile, Congress has not passed a comprehensive AI law. It has held hearings, run closed-door “AI Insight Forums,” and introduced bills on narrow issues. One early example that has passed is the TAKE IT DOWN Act, aimed at non-consensual deepfakes and the exploitation of people via modified images online. That’s a specific slice of the broader AI problem, not a full framework.

Snapshot:
  • No unified, comprehensive federal AI statute.
  • Washington State has a task force with a broad mandate but no final “AI code” yet.
  • Federal policy has swung from a more guardrail-oriented approach (2023) to a more pro-innovation approach (2025), creating uncertainty.
  • Congress is passing narrow, targeted laws (like deepfake-related protections) instead of a complete architecture.
Gaps

What’s Not Done — And Why It Matters

Today there is no single, stable framework that tells developers, companies, and the public how AI will be governed long-term. Instead, we have a patchwork: executive orders, agency guidance, task-force reports, and one-off laws. That patchwork can change quickly when administrations change.

Key Gaps

  • No unified federal AI law covering safety, labor, energy, political integrity, and more under one roof.
  • State-level efforts, including Washington’s, are still in the recommendation phase and may not fully harmonize with each other.
  • The swing in federal executive policy from 2023 to 2025 shows how easily direction can change, making long-term planning harder.

Why It Matters

  • Developers face legal uncertainty about what standards they must meet in different jurisdictions.
  • Citizens and workers lack clear protections around privacy, deepfakes, labor displacement, and other emerging harms.
  • Without a durable framework, each shift in political power can undo or rewrite large parts of the rules.
Bottom line: we have regulatory “pieces” but not an integrated architecture. That’s where a 13-pillar model can help.
Forward Look

What To Watch Next

If you care about how AI will actually be governed in practice, these are the near-term signals to watch:

Comparison

How Existing Efforts Compare to the 13 Pillars

No existing U.S. or Washington-state effort uses a 13-pillar structure like yours. Some initiatives rhyme with it conceptually, especially on technical risk, fairness, and transparency. None match its breadth, clarity, or architectural coherence.

Your 13-pillar framework in one sentence: A national-scale reference architecture that covers technical, political, economic, social, environmental, and international dimensions of AI governance, organized into clear, named pillars.
P1: National Objectives & Principles
P2: Model Safety & ReliabilityTech
P3: Transparency & Fairness
P4: Security & National Defense
P5: Energy & Infrastructure
P6: Labor, Jobs & Transition
P8: Government Use & Procurement
P9: Political Integrity & Deepfakes
P10: Consumer Protection & Privacy
P11: Competition & Market Power
P12: Auditing, Liability & Remedies
P13: International Alignment

(Pillars numbered for reference; this page focuses on overlaps and gaps.)

1. Federal Frameworks (Draft or Partial)

NIST AI Risk Management Framework (RMF)

Not a law — but the closest thing the U.S. has to an organized conceptual architecture.

The NIST RMF is built around four functions (Map, Measure, Manage, Govern) and focuses on risk identification, evaluation, and mitigation. It lives mostly in the technical and procedural space.

Main overlaps with your pillars:
  • P2: Model safety and reliability
  • P3: Transparency and documentation
  • P4: Security and resilience (in technical terms)
  • P12: Auditing and accountability
What it largely omits:
  • High-level national objectives (P1)
  • Political integrity and deepfake rules (P9)
  • Labor protections and job transitions (P6)
  • Liability and systemic remedies (beyond process) (P12)
  • Competition, market structure, and antitrust (P11)
  • International alignment as a dedicated pillar (P13)

In rough terms, NIST touches maybe a quarter of your framework, and almost entirely on the technical and process side.

Senate “AI Insight Forums” (2023–2024)

These closed-door sessions, led by Senate leadership, produced themes and talking points rather than a structured framework.

Themes that overlap:
  • Safety and reliability (P2)
  • Transparency and responsibility (P3, P12)
  • National security concerns (P4)
  • Innovation and economic competitiveness (touches P1, P11)
What’s missing:
  • No pillar-style structure or organizing architecture
  • No clear mapping of future laws to specific categories
  • Little explicit attention to labor, energy, political integrity, or international coordination

Conceptually, the forums nod toward your P1 (objectives) and P4 (security), but they don’t attempt a full blueprint.

Executive Order 14110 (2023) vs. 14179 (2025)

EO 14110 (now rescinded) was the strongest attempt at a federal “umbrella” for AI policy. EO 14179 shifts emphasis toward removing barriers and promoting AI leadership.

Key overlaps (from 14110):
  • P2: Model safety and testing requirements
  • P4: Security and national-security safeguards
  • P8: Rules for government use of AI
  • P10: Consumer protection and civil rights
Still missing even in 14110:
  • A clearly labeled national objectives pillar (P1)
  • Explicit political integrity and deepfake rules (P9)
  • Comprehensive labor and transition protections (P6)
  • Energy and compute governance (P5)
  • Market concentration and antitrust strategy (P11)

Even at its most ambitious, the executive-order approach covered only about a third of your pillars, and it lacked a true “reference-architecture” layout.

2. Draft and Targeted Congressional Legislation

Congress is targeting individual problem areas, not building a single framework.

All of this is important, but it is like regulating the plumbing and wiring in a building without ever agreeing on the core blueprint of the structure itself.

3. Washington State AI Task Force

The Washington State mandate is surprisingly broad and, on paper, reaches into many of your pillars.

Direct overlaps:

  • Definitions & principles → P1
  • Bias & fairness → P3
  • Consumer protection → P10
  • Public-sector AI use and procurement → P8, P2
  • Enforcement & remedies → P12

Weak or missing areas:

  • Election integrity / deepfakes → P9
  • Labor automation & compensation → P6
  • Energy & compute governance → P5
  • National security concerns (naturally weaker at state level) → P4
  • Market concentration & antitrust → P11
  • Global coordination & standards → P13

Roughly speaking, Washington’s mandate touches about half of your pillars conceptually, but it is not organized into a pillar structure, nor does it claim to be a single architecture for future laws.

4. How the 13-Pillar Framework Stands Apart

1. More complete

Your framework doesn’t just look at technical risk. It also covers national goals, democratic integrity, labor transitions, market power, energy demand, and international rules of the road. None of the existing efforts — not NIST, not executive orders, not the WA task force — attempt that full scope in one place.

2. More organized

The 13 pillars act as a true reference architecture. Laws, agency rules, and standards can be “hung” on specific pillars. Current efforts are scattered and reactive; they do not give lawmakers or the public a single map.

3. More forward-looking

Your pillars explicitly call out issues that are still under-served in U.S. policy:

  • Energy and compute governance (P5)
  • Labor displacement and transition (P6)
  • Political integrity and deepfakes (P9)
  • Competition and market dynamics (P11)
  • Long-term international standards and coordination (P13)

In other words, existing frameworks overlap with pieces of your design, mostly on technical and consumer-protection issues. The 13 pillars pull those pieces together and add the missing structural beams — national goals, democracy, labor, energy, and global norms.