Reference shelf

Recurring AI reports and indexes worth checking before repeating the internet's loudest takes.

This page is for slower, sturdier sources. Instead of another rolling news stream, it collects the recurring reports, annual indexes, and benchmark dashboards that help anchor AI coverage in something more durable.

8

Tracked report sources

Annual + ongoing

Cadence mix

Market + labour + policy

Coverage focus

Stanford HAI

AI Index

Annual

Focus

Industry, research, investment, policy, labour, and education

Why it matters

One of the most useful broad reality-check reports because it combines technical progress with economic and labour data.

Best for

Big-picture market context and grounded charts

Open source report

McKinsey

The State of AI

Annual

Focus

Enterprise adoption, business value, deployment patterns, and executive sentiment

Why it matters

Useful for separating public AI hype from what larger organisations are actually deploying.

Best for

Enterprise adoption and operating-model signals

Open source report

PwC

AI Jobs Barometer

Annual

Focus

Labour demand, skills, productivity, and wage signals linked to AI

Why it matters

One of the clearest recurring labour-market views for checking whether AI hiring claims line up with real job-market data.

Best for

Employment and skills trends

Open source report

Air Street Capital

State of AI Report

Annual

Focus

Frontier-model landscape, funding, open-source momentum, and strategic market shifts

Why it matters

Strong at synthesising market structure and competitive movement across labs, startups, and open models.

Best for

Strategic market snapshots and slide-friendly charts

Open source report

Epoch AI

Notable AI Models

Continuously updated

Focus

Model timelines, compute, parameter counts, and historical trend data

Why it matters

A useful factual backbone for tracking long-run model growth without relying on social-media summaries.

Best for

Model-history charts and technical scaling context

Open source report

OECD

OECD AI Policy Observatory

Ongoing

Focus

Policy tracking, national AI initiatives, and public-sector measures

Why it matters

Helpful when the story is regulation, national strategy, or policy movement rather than model launches.

Best for

Policy and government tracking

Open source report

Stanford CRFM

HELM Benchmarking

Ongoing

Focus

Holistic model evaluation across scenarios, tradeoffs, and benchmark methodology

Why it matters

Important when a reader wants nuance around evaluation quality rather than just a single headline score.

Best for

Benchmark methodology and model-quality context

Open source report

Artificial Analysis

Artificial Analysis

Frequent

Focus

Model quality, speed, price, and arena-style comparisons

Why it matters

Closer to an operating dashboard than a yearly report, and very useful for keeping current comparisons grounded.

Best for

Fast-moving model and provider comparisons

Open source report