AI-Regulation Intensity

What this driver is

AI regulation intensity is a continuing meter of binding rules, designations, and enforcement aimed at general-purpose and high-risk AI systems. It is not the NEW event bundle EU platform regulation (DSA/DMA). DSA/DMA are intermediary and gatekeeper statutes. It is not the NEW ai-race bundle. This node is how hard the law actually bites on AI systems.

Those neighbouring nodes can move with this meter. They do not replace it.

It applies globally as a time-varying intensity. The live Factrail series uses World Bank Regulatory Quality (GOV_WGI_RQ_SC) as a disclosed proxy for 2020–2024, the same series attached to AI Safety Governance. Stored values ran from 0.54307721 in 2020 to 0.54417717 in 2024. Sharing a proxy is not sharing an identity. Missing 2025 releases stay missing. This page does not interpolate. General regulatory quality is not the AI Act.

The documentary neighbourhood includes Regulation (EU) 2024/1689, EO 14110, and national AI statutes. Ursula von der Leyen’s Commission and Margrethe Vestager’s digital portfolio are live people on the Union side. They are referenced, not reminted.

Mechanism

Regulation intensity moves welfare through named channels.

First, statutes. The AI Act entered into force on 1 August 2024 after adoption on 13 June 2024, with staged application. Prohibited-practice and GPAI duties are the Union overlay. This page does not invent a unique compliance-hour integer.

Second, designations and codes. High-risk lists, GPAI codes of practice, and national implementations convert a regulation into operator duties. A code is not a completed court judgment.

Third, enforcement. Fines, market-surveillance, and extra-territorial reach change expected cost. Contested euro amounts stay on named Commission decisions. This page does not invent Contribution.size.

Fourth, the proxy’s limits. WGI Regulatory Quality is a broad perception score. The same series also meters AI Safety Governance; the two questions remain different.

Timing is months for an application clock and years for a WGI point. A later dedicated AI-statute coverage series would be a better meter.

Evidence base

The live quantitative record is Regulatory Quality — Governance score (0–100) (https://api.worldbank.org/v2/country/all/indicator/GOV_WGI_RQ_SC?date=2020:2025&format=json&per_page=20000 (opens in a new tab)). Context sits at EUR-Lex for Regulation (EU) 2024/1689 (https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32024R1689 (opens in a new tab)) and at the Federal Register for EO 14110.

This page does not invent a unique fine total, model-count, or Contribution.size. Where WGI prints a score, that score stays on WGI. Where the Official Journal prints an application date, that date stays on the Journal.

Evidence is strongest for the stored proxy observations and for the named AI Act. It is weaker for the claim that every RQ movement was AI law. It is weakest where a “Brussels vs innovation” talking point is asked to replace the series.

Effect on welfare indicators

This driver, when it actually runs, is a first-order channel into civic and consumer protection around automated decisions.

• Having a say in government: direction mixed-to-higher when transparency, prohibited-practice bans, and contestable duties actually apply; direction lower if rules become an unaccountable prior restraint. Horizon: months to years after application clocks. Confidence: medium. The indicator is higher-is-better. Strengthening protective AI regulation is treated here as welfare-positive unless a sourced speech-restrictive application dominates. No unique fine integer is invented here.

Missing coefficients stay missing.

The welfare sign of a completed protective AI statute is not automatically negative. A slogan does not rebate the Official Journal.

Interactions and contested points

Reinforcement is obvious. Offsetting is the claim that AI law merely duplicates DSA. The AI Act is a different instrument.

The strongest competing view is that WGI Regulatory Quality cannot carry AI regulation, and that this node should wait for a dedicated series. That objection is serious. The competing defence is the existing disclosed proxy plus the named 2024 Act as neighbourhood, not as a unique Factrail coefficient.

A second competing view is that this node is the same thing as ai-safety-governance or platform-regulation-eu. It is not. Safety governance is evaluation practice. DSA/DMA are platform statutes. This meter is AI-specific legal intensity.

Not established: a unique fine or model integer as Contribution.size. A later dedicated series would update the meter.