AI Frontier Monitor — W/E 17 April 2026

Ramparts
AI Frontier Monitor
Asymmetric Intelligence
AI Intelligence Brief ● Live

Ramparts AI Frontier Monitor

Week of W/E 17 April 2026 · Asymmetric Intelligence · Published 2026-04-17T09:00:00Z T09:00:00Z

Issue
Vol. 1 · Issue 4
Jurisdiction Grid
12 countries
↓ Print / PDF
↯ This Week’s Five Key Changes
  • → Anthropic dual opsec failure: KAIROS/BUDDY/undercover mode + Mythos leak M00
  • → GSA ‘any lawful use’ clause: comment deadline April 3 — generalises Pentagon Anthropic playbook to all federal AI M09 / M10
  • → OpenAI $852B valuation at 60x revenue — most extreme in tech history M03
  • → AISI pipeline: all three principal AI safety bodies in leadership transition approaching August 2026 enforcement cliff M15 / M10
  • → EU AI Act Omnibus: 28 April trilogue agreement target — Standards Vacuum 122 days to deadline M09
00

The Signal

⚡ The Signal — Week of W/E 17 April 2026

No harmonised standards published in Official Journal as August 2026 high-risk compliance deadline approaches, confirmed by EU AI Office and independent trackers. CEN-CENELEC drafts in progress but vacuum creates compliance uncertainty for providers. Governance velocity lags capability deployment in regulated sectors; watch for delegated acts or emergency standardisation.

01

Executive Insight

No items this issue.

02

Model Frontier

Benchmark Leaderboard — W/E 17 April 2026

ARC-AGI-2 (Static Reasoning)

Human average ~60%. Models above this line are superhuman on this benchmark.

ModelScore

ARC-AGI-3 (Interactive/Agentic)

Humans score 100%. Frontier AI near-zero — reveals the adaptive intelligence gap.

ModelScore

GPQA Diamond (Graduate Science)

Human expert ceiling ~70–80%.

ModelScore
03

Investment & M&A

No items this issue.

04

Sector Penetration

No items this issue.

05

European & China Watch

🇪🇺 European AI

No items this issue.

Funding Rounds >$50M

No items this issue.

Incumbent Displacement

No items this issue.

🇨🇳 China AI

No items this issue.

DeepSeek

No items this issue.

Alibaba

No items this issue.

Export Controls

No items this issue.

06

AI in Science

No items this issue.

07

Risk Indicators: 2028

Governance Fragmentation

HIGH

EU/US governance schism: EP Omnibus vote (569–45–23) vs White House nonbinding framework explicitly opposing new federal AI regulator. Domestic US patchwork (Colorado, New York, Texas, Georgia, Minnesota) cannot yet be federally preempted. Labs have structural incentive to route EU compliance through US legal entities.

↗ Source

Cyber Escalation

HIGH

AI agent orchestration infrastructure (Langflow, LiteLLM, n8n) is established attack surface. CVE-2026-33017 (CVSS 9.3, Langflow) exploited within 20 hours of advisory — no public PoC required. CVE-2026-33634 (Trivy/LiteLLM supply chain) found in 36% of monitored cloud environments. CISA enforcement only applies to federal agencies; private sector exposure unaddressed.

↗ Source

Platform Power

ELEVATED

EU AI Act Omnibus implementation delays structurally benefit incumbent labs. SpaceX–xAI merger (Feb 2026) created entity combining Grok LLM + Starlink + DoD/NASA contracts + X social platform — no single regulator has comprehensive oversight. Hyperscaler capex approaching $750B in 2026.

↗ Source

Export Controls

ELEVATED

US Commerce Dept withdrew planned AI chip export rule (13 March 2026). No replacement rule published. Biden-era AI Diffusion Rule rescinded May 2025. 12–18 month enforcement vacuum for model weight exports. Cloud IaaS providers can currently serve Chinese customers via foreign data centers without triggering hardware export controls. Draft new regulations (5 March) not yet published as formal text.

↗ Source

Disinfo Velocity

HIGH

EEAS 4th Annual FIMI Report: 1 in 4 detected incidents in 2025 involved AI tools. AI deepfakes confirmed in 2026 US midterm campaigns (NRSC, Reuters 28 March). No federal law restricts AI in political advertising. Russia targeting Armenia ahead of June 2026 elections. FIMI/financial fraud convergence documented by Meta H1 2026 ATR.

↗ Source

Standards Vacuum

HIGH

EU AI Act high-risk obligations approach without harmonised standards

⚡ 2028 Horizon Signal

Creates definitional arbitrage for GPAI models

↗ Source
08

Military AI Watch

No items this issue.

09

Law & Guidance

EU AI Act — Layered System

Standards vacuum active · 107 days to general application

Layer 1 — AI Act Text AI Act Text Active

Timeline: Unchanged since 2026-03-30

No items this issue.

Layer 2 — Delegated / Implementing Acts Delegated / Implementing Acts In progress

Timeline: Unchanged since 2026-03-30

No items this issue.

Layer 3 — Harmonised Standards (CEN-CENELEC JTC21) Harmonised Standards (CEN-CENELEC JTC21) Delayed

Timeline: Unchanged since 2026-03-30

No items this issue.

Layer 4 — GPAI Code of Practice GPAI Code of Practice In progress

Timeline: Unchanged since 2026-03-30

No items this issue.

Layer 5 — National Enforcement (NCAs) National Enforcement (NCAs) Not yet active

Timeline: Unchanged since 2026-03-30

No items this issue.

Layer 6 — AI Office Supervisory Decisions AI Office Supervisory Decisions Not yet active

Timeline: Unchanged since 2026-03-30

No items this issue.

Layer 7 — Digital Omnibus Trilogue Digital Omnibus Trilogue Active — in trilogue

Timeline: Unchanged since 2026-04-01

No items this issue.

Country Grid — Law & Standards Status

🟢 Binding law in force  ·  🟡 Law passed/in implementation  ·  🟠 Guidance/soft law only  ·  ⚪ No framework  ·  🆕 New this week  ·  ⚠️ Amendment/enforcement

Jurisdiction Binding Law Key Guidance Last Updated 🔔
EU 🟡 AI Act (GPAI Aug 2025; high-risk delayed to Dec 2027/Aug 2028 pending Omnibus) 2026-04-17 —
USA 🟠 No federal law; state laws: CO ADMT (Jan 2027 if rewrite passes), TX TRAIGA, WA HB 2225/HB 1170 (signed), NY 20 Mar 2026 —
UK 🟠 No single statute; Data Use & Access Act 2025 (TDM provisions); ASGARD contracts operationalising AI targeting 2026-04-17 —
China 🟢 GenAI Measures 2023; AI Content Labelling Sep 2025; CSL Jan 2026 (AI governance incorporated); Embodied AI Standard (effective Jun 1, 2026) 2026-04-17 —
India 🟠 IT Rules 2026 (SGI, Feb 20) 2026-04-17 —
Canada 🟠 AIDA dead; new AI strategy in development Jan 2025 —
Brazil 🟡 PL 2338 in Chamber of Deputies (not yet law) Dec 2024 —
South Korea 🟢 AI Basic Act in force Jan 22, 2026 Jan 2026 —
Japan 🟢 AI Promotion Act fully effective Sep 2025 Sep 2025 —
Singapore 🟠 Model AI Governance Framework (voluntary) Mar 2026 —
Australia 🟠 Voluntary framework; mandatory in development 22 Mar 2026 —
UAE 🟢 UAE AI Act 2026 — in force Mar 2026 (grace to Sep 2026) Mar 2026 —
US OpenAI $852B close (31 March); GSA ‘any lawful’ clause (April 3 deadline); White House AI Framework preemption call; Anthropic Pentagon injunction granted; NBER 9x job-cut acceleration projection. 2026-04-17 —
Germany German Army committed to deploying AI for battlefield analysis using Ukrainian combat data; Maven cited as reference architecture. 2026-04-17 —
Saudi Arabia Prosperity7 (Aramco) invested in Gimlet Labs (AI agent infrastructure) — sovereign commodity wealth hedging against AI disruption of energy sector by owning infrastructure layer. 2026-04-17 —

Country Watch — Threshold Tracker

Countries approaching entry to the grid.

JurisdictionThreshold & TriggerCurrent Framework
Armenia ↗
Hungary ↗
10

AI Governance

11

Ethics & Accountability

No items this issue.

12

Technical Standards

No items this issue.

13

Litigation Tracker

No items this issue.

14

Personnel & Org Watch

Lab & Industry Movements

Weekly Digest

The signal in your inbox every Thursday

Primary sources only. No press summaries. 09:00 GMT without exception — reliable enough to build into your morning routine.

No spam. Unsubscribe in one click. Published by Ramparts, Gibraltar.

Weekly Digest

The signal in your inbox every Thursday

Primary sources only. No press summaries. 09:00 GMT without exception — reliable enough to build into your morning routine.

No spam. Unsubscribe in one click. Published by Ramparts, Gibraltar.