← All pages

METR

An organization in the news record
KindAn organization · graph entity E2 · Wikidata Q135185153
Wikipedia's sources0 of the 0 references dated 2026-09-01 to 2026-11-21 in METR · 3 more from the year before the record
Views72 quoted, from 5 source domains
Dates2026-09-23 to 2026-11-20 · 6 weeks
Arguments10 · AI accelerating AI R&D, Near-term AI R&D automation, Alignment eval validity and eval awareness, Scope of AI developer risk reports, AI accelerating AI and science, AI incidents and loss of control …
Reported byDwarkesh Patel, Latent Space, METR, Transformer, Zvi Mowshowitz
Profilea readable profile, written from these views and checked sentence by sentence
Only what the organization says as itself. Views of its staff, teams or founders, named or not, are not counted here.

Where they stand

By argument, from their own words. ✓ agree · ✗ disagree · ~ no clear side · · side unclear (the two readings differ).

AI incidents and loss of control ~

“Recent AI incidents show frontier AI is not under adequate control and labs should slow down or face liability”

AI accelerating AI R&D ✓ ~

“AI is substantially accelerating the pace of AI capabilities research and may accelerate it further.”

Near-term AI R&D automation ✓ ~ ·

“AI will automate AI research and reach recursive self-improvement within the next few years (by about 2027-2028)”

Alignment eval validity and eval awareness ✓ ~ ·

“AI models can often tell alignment evals apart from real tasks, which undermines the validity of those evals”

Scope of AI developer risk reports ·

“Frontier lab risk reports should cover safety-relevant decision-making undermining and models still in training”

AI accelerating AI and science ✓

“AI is substantially accelerating AI development and scientific progress”

Containing AI with classical security ~

“Classical cybersecurity measures can adequately contain AI systems and defend organizations against AI-related risk”

Automated R&D catastrophic risk ✓

“Direct catastrophic risk from automated AI R&D in current models is low.”

benchmark saturation ✗

“This benchmark is unlikely to get saturated since the bar is set by other smart models”

reward hacking effort in evaluation development ·

“Anti-reward-hacking effort will need to increase as agents find more niche exploits.”

Said differently

None in the record.

History

Built 2026-11-21 06:50 UTC from the news knowledge graph by code. Every quote was checked against the source's own words; which side a quote takes is a model's reading, made twice.