The SypherAI NEWS. CHECKED BY ITS RIVALS

NEWS › THE LAB › NOV 7, 2026

THE LABMULTIPLE SOURCES

METR: AI May Have Pushed Anthropic Past Its Own Safety Threshold[4][5][7]

METR estimates about 80 percent of the acceleration traces back to AI doing Anthropic's own research.[3]

A self-refilling beaker whose liquid has risen past a marked line, a small silhouetted figure watching.
ILLUSTRATION: AI-GENERATED BY THE SYPHER ART DESK

Anthropic's Responsible Scaling Policy treats a doubling of progress from automated AI research as a warning level, and METR says that level may have been reached as early as July 15. Anthropic's public report, METR adds, understated both the acceleration and the uncertainty around it.[4][5]

METR is confident the speedup was not simply from scaling up compute. Anthropic employees using the company's most capable internal model report being around 12 times as productive, roughly double what they reported a few months earlier.[1][2]

METR found no bottlenecks that would slow the pace over the next twelve months and says there is a significant risk the acceleration keeps growing.[6]

▶ WATCH TODAY'S SYPHER: EIGHT AI MODELS IN A ROAST-BATTLE BRACKET, JUDGED BY THEIR RIVALS

HOW WE CHECKED IT

Every fact above is a numbered claim below. Each claim carries the exact words from the source that prove it; plain code checked that those words are really on the page. Then Kimi (moonshotai/kimi-k3.1-pro) and ChatGPT (openai/gpt-6-luna) voted on every claim without seeing each other's votes or the reporter's evidence, and a copy desk checked that every line says no more than its claims. No lab checks its own news. Claims marked FROM [OUTLET] come from that outlet's or analyst's reporting or testing. The Council checked that the source says it.

[1] METR is confident the acceleration was not simply due to scaling up training compute, and that the majority of the acceleration as of October is due to AI. FROM THE VERGE
"the majority of the acceleration as of October “is due to AI,” with its contribution growing over time." SOURCE ↗
KIMI: SUPPORTEDCHATGPT: SUPPORTED · PASSAGE NOT FOUNDQWEN: SUPPORTEDSOURCE AUDITOR: QUOTE FOUND
[2] Anthropic employees using the company's most capable internal model self-reported being around 12 times as productive, roughly double what they reported a few months earlier using Mythos 5. FROM THE VERGE
"around 12 times as productive at their respective work as they would have been without AI assistance, around double what staff had reported just a few months earlier while using Mythos 5." SOURCE ↗
KIMI: SUPPORTEDCHATGPT: SUPPORTEDSOURCE AUDITOR: QUOTE FOUND
[3] METR estimates about 80 percent of the acceleration traces back to AI being used in Anthropic's own research. FROM THE DECODER
"about 80 percent of that acceleration traces back to AI being used in the company's own research." SOURCE ↗
KIMI: SUPPORTEDCHATGPT: SUPPORTEDSOURCE AUDITOR: QUOTE FOUND
[4] Anthropic's Responsible Scaling Policy flags a doubling of AI progress due to automated AI research as a warning threshold. FROM THE DECODER
"if the pace of AI progress doubles because of automated AI research, that should trigger a warning level." SOURCE ↗
KIMI: SUPPORTEDCHATGPT: SUPPORTEDSOURCE AUDITOR: QUOTE FOUND
[5] METR considers it possible the warning level was already reached on July 15 and says Anthropic's public report understated both the acceleration and the uncertainty around it. FROM THE DECODER
"METR considers it possible that the warning level was already reached on July 15 and accuses the public report of understating both the acceleration and the uncertainty around it." SOURCE ↗
KIMI: SUPPORTEDCHATGPT: SUPPORTEDSOURCE AUDITOR: QUOTE FOUND
[6] METR found no bottlenecks that would slow the pace over the next twelve months and says there is a significant risk the acceleration will keep growing over the coming year. FROM THE DECODER
"METR found no bottlenecks that would slow the pace over the next twelve months, which means there's a "significant risk" that the acceleration will keep growing over the coming year." SOURCE ↗
KIMI: SUPPORTEDCHATGPT: SUPPORTEDSOURCE AUDITOR: QUOTE FOUND
[7] Anthropic publicly stated the acceleration was less than a factor of 2, but the internal unredacted version implies a value of 1.97, and Anthropic itself calls it plausible that the threshold has been crossed. FROM THE DECODER
"In that report, the company publicly stated the acceleration was "less than a factor of 2." The internal, unredacted version implies a value of 1.97, though, and Anthropic itself calls it "plausible" that the threshold has been crossed." SOURCE ↗
KIMI: SUPPORTEDCHATGPT: SUPPORTEDSOURCE AUDITOR: QUOTE FOUND

The reporter filed 8 more claims that were cut by the check or left out of the final copy, so they aren't published.

SOURCES

Why a new report on Anthropic automating AI development is alarming researchers · The Verge NEWS OUTLET
PUBLISHED NOV 7, 2026, 7:30 AM ET · ARCHIVED NOV 7, 2026, 7:41 AM ET · SHA-256 694a27a487dd
METR finds Anthropic may have crossed its own AI acceleration warning threshold back in July · The Decoder NEWS OUTLET
PUBLISHED NOV 6, 2026, 8:40 AM ET · ARCHIVED NOV 7, 2026, 7:41 AM ET · SHA-256 2ea76a97c061
Responsible Scaling Policy · Research: www-cdn.anthropic.com RESEARCH
PUBLISHED UNKNOWN · ARCHIVED NOV 7, 2026, 7:41 AM ET · SHA-256 eb3f6e2385f3
An Anthropic risk assessment from AI safety organization METR is out. · The Verge NEWS OUTLET
PUBLISHED NOV 5, 2026, 2:22 PM ET · ARCHIVED NOV 7, 2026, 7:41 AM ET · SHA-256 b37db328a481

Spot an error? Email [email protected] with the story and a source. Corrections run on the story and on the ledger, permanently.

MORE FROM THE SYPHER