Washington Examiner puts ‘independent’ in quote marks; METR says ‘our funders have no say in the projects we work on’

The claim. The Washington Examiner reports that Tarbell Center and METR, two nonprofits warning about the dangers of AI, are "backed by allies of the technology firm Anthropic and its regulatory agenda". It names Dustin Moskovitz, Jaan Tallinn and Paul Christiano, and credits its own review of public statements. No other report we hold mentions who funds either group.

The denial is in the same report. A Tarbell spokesperson says "our funders have no editorial role in the journalism we support". METR president Chris Painter says "our funders have no say in the projects we work on". The critics of the two groups are quoted in that report but not named.

On the policy itself, five ways:

Outlet	What it says	Sourced
Washington Examiner	the US "would be insane" to pause	no source named
National Review	Congress must tell "a guardrail" from "a roadblock"	no source named
The New York Times	much of the regulatory effort "has stalled"	its own reporter
The Washington Post	Trump is on "the unpopular side"	no poll figure
CNBC	Palantir's Alex Karp wants "reasonable guidelines"	one executive

Not settled. Whether the funders shape what the two groups say. The Examiner frames them as "'independent' groups"; both deny donor control; no report we hold tests either account. We hold none naming who pays for the opposite case, though three reports make it. And none gives a number for what the public thinks.

This is an argument about who is paying the organisations that tell Congress artificial intelligence is dangerous.

How it got here. Anthropic is an American AI company, founded in 2021, and one of the firms building the systems at issue. The Washington Examiner reports that Dustin Moskovitz and Jaan Tallinn — it describes them as a Meta cofounder and a Skype cofounder — were early investors shortly after that founding; that Paul Christiano was an initial trustee of Anthropic's Long-Term Benefit Trust in 2023 and founded the Alignment Research Center; and that he later left the trust for the Anthropic-backed AI Safety Institute, an advisory body to the UK government. The same report says Anthropic chief executive Dario Amodei "raised eyebrows" this month with an essay promoting METR as a "third-party" evaluator of his own company's tools, and urging every major AI company to take the same step.

Who is in it. Congress would have to write any federal AI law. Tarbell Center runs a fellowship placing technology reporters at established news outlets. METR — Model Evaluation and Threat Research — runs risk evaluations; its website says nine AI developers have adopted its "Responsible Scaling Policies". Palantir, whose chief executive Alex Karp asked for "reasonable guidelines", sells data systems to governments. The Washington Examiner writes that President Donald Trump "sees the bigger picture" in refusing to slow development; The Washington Post writes that he has put himself on the unpopular side of it.

What is at stake. If the warnings are independent, they are evidence for rules on the most powerful systems. If they are an interested party's advocacy, the same warnings become lobbying by one company's backers.

Where it stands. No federal AI law. The New York Times says much of the effort "has stalled"; The Washington Post says Democrats eyeing the presidency are "heavier on rhetoric than specifics".

What is being argued about

Who funds Tarbell Center and METR, and whether that makes them independent
Whether Congress should slow frontier AI development, or race China
Whether AI regulation is popular, and with whom
Why no federal AI law has passed

Five outlets, one week, five different argumentsSide by side

Outlet	What it says the argument is	Named source for it
Washington Examiner	Two "'independent'" groups warning about AI are "backed by allies of the technology firm Anthropic"	Names three funders; quotes both groups denying donor control
Washington Examiner (column)	The US "would be insane" to pause while China does not	None; the pause's advocates are "performing for public opinion"
National Review	Congress needs "to distinguish between a guardrail and a roadblock"	None in the text we hold
The New York Times	Regulating AI is hard and "much of that effort has stalled"	Its own technology reporter, Cecilia Kang
The Washington Post	Trump has put himself "on the unpopular side" of the AI debate	No poll figure or method given
CNBC	The industry needs "reasonable guidelines"	Palantir chief executive Alex Karp

What each report says the AI argument is about, in its own words. Four of the six below published only a headline and an opening line, which is all we hold of them.

Sources

Washington Examiner

National Review

The New York Times T The Washington Post

CNBC

Claims and verdicts

unverified The Tarbell Center and METR, which publish warnings about AI risk, are funded by people who backed Anthropic — Dustin Moskovitz, Jaan Tallinn and Paul Christiano.

Only the Washington Examiner reports this, from what it calls its own review of public statements, and no other report we hold addresses the funding of either group. Neither organisation denies the funders in the quotes the Examiner publishes.

Checked against 1 source:

Washington Examiner

disputed Anthropic's allies shape what those groups say about AI.

The Washington Examiner frames the two groups as "'independent' groups" backed by Anthropic allies, while a Tarbell spokesperson tells the same paper that "our funders have no editorial role" and METR president Chris Painter says "our funders have no say in the projects we work on." We hold no report testing either side's account.

Checked against 1 source:

Washington Examiner

true Efforts in Congress to regulate AI have largely stalled.

The New York Times says much of the regulatory effort has stalled, and The Washington Post reports Democrats eyeing the presidency are "heavier on rhetoric than specifics" on AI. National Review and the Washington Examiner describe an argument still being had rather than a law being passed.

Checked against 4 sources:

The New York Times T The Washington Post

National Review

Washington Examiner

Who wrote it

H Hudson Crozier Washington Examiner

What the bias passes found in their reporting

Quotation marks used to doubt the groups' independence is put in scare quotes in the headline; both groups' denials are quoted at length in the body×1

Right of reply both accused organisations are quoted in their own words, one of them twice×1

Scare quotes in the headline the groups' independence is marked as doubtful before any evidence is given×1

Unnamed accusers the central criticism is attributed to "critics" with no one named×1

Claims we checked that rest on their reporting

1 disputed1 unverified

From 46 reports we hold under this byline, at Washington Examiner. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

T Tom Rogan Washington Examiner

What the bias passes found in their reporting

Contested characterisation stated flat a genocide finding is asserted in the writer's own voice, unattributed, as a rhetorical question×1

Contested claim stated flat a genocide finding is asserted inside a rhetorical question, unattributed×1

Speculation written in the indicative a hypothetical future is given in "it would mean", with no source and no hedge×1

Claims we checked that rest on their reporting

1 true

From 28 reports we hold under this byline, at Washington Examiner. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

C Cecilia Kang The New York Times

What the bias passes found in their reporting

Company's knowledge says the company learned late; the other reports do not×1

Verb chosen 'Meddled' stated flat; success is implied×1

Verb chosen 'meddled', stated flat, with no mention of whether any attempt succeeded×1

Who found it discoverer unnamed; the company is cast as learning late×1

Claims we checked that rest on their reporting

1 disputed1 unverified1 true

From 13 reports we hold under this byline, at The New York Times, The Sydney Morning Herald. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

J June Kim The New York Times

No independent reading has coded a report under this byline yet. We do not infer a journalist's views from the outlet that employs them.

Claims we checked that rest on their reporting

1 true

From 2 reports we hold under this byline, at The New York Times. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

N Nikolay Nikolov The New York Times

No independent reading has coded a report under this byline yet. We do not infer a journalist's views from the outlet that employs them.

Claims we checked that rest on their reporting

1 true

From 7 reports we hold under this byline, at The New York Times. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

N Nour Idriss The New York Times

No independent reading has coded a report under this byline yet. We do not infer a journalist's views from the outlet that employs them.

Claims we checked that rest on their reporting

1 true

From 6 reports we hold under this byline, at The New York Times. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

K Kat Cammack National Review

What the bias passes found in their reporting

Instruction to lawmakers the standfirst tells the legislature what it needs rather than what it did×1

Claims we checked that rest on their reporting

1 true

From 1 report we hold under this byline, at National Review. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

L Liz Goodwin The Washington Post

No independent reading has coded a report under this byline yet. We do not infer a journalist's views from the outlet that employs them.

Claims we checked that rest on their reporting

1 true

From 10 reports we hold under this byline, at The Washington Post. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

D Dan Merica The Washington Post

What the bias passes found in their reporting

Judgement in the lede the president's position is called unpopular with no polling figure or method attached×1

Popularity claimed without a number which side is "unpopular" is asserted; no poll, method or figure is given×1

From 43 reports we hold under this byline, at The Washington Post. Two independent readings, each made without the masthead, a quote attached to every tag. A record, never a rating. Their full record →

The bylines on the reports this story is built from. Hover one for what our passes have measured on their reporting. · 1 report we hold carries no byline — we do not invent one.

Coverage, outlet by outlet

CNBC — independent

AI needs to have 'reasonable guidelines,' Palantir's Karp tells CNBC

The only industry voice quoted by name on what the rules should look like: Palantir's Alex Karp asking for "reasonable guidelines".

Palantir CEO Alex Karp called for "reasonable guidelines" on artificial intelligence regulation as tech and policy leaders deliberate over the technology.

Whose voice frames the regulation: the headline is one defence contractor chief executive's phrase, with no critic quoted — “Palantir CEO Alex Karp called for "reasonable guidelines" on artificial intelligence regulation as tech and policy leaders deliberate over the technology.”

Read it at CNBC → https://www.cnbc.com/2026/09/17/ai-safety-palantir-karp.html

Washington Examiner — independent

Anthropic allies are behind ‘independent’ groups ringing alarm bells about AI

The only report naming the funders behind Tarbell Center and METR — Dustin Moskovitz, Jaan Tallinn, Paul Christiano — and the only one carrying both organisations' denials in their own words.

A pair of organizations issuing warnings about the possible dangers of artificial intelligence are backed by allies of the technology firm Anthropic and its regulatory agenda. The nonprofit groups Tarbell Center and Model Evaluation and Threat Research have emerged as key players in the public debate surrounding AI risks and potential policy, promoting AI-skeptical journalism and risk evaluations, respectively. The Tarbell Center has ties to Meta cofounder and longtime pro-Democrat donor Dustin Moskovitz and Skype cofounder Jaan Tallinn, while METR is connected to Moskovitz, Tallinn, and AI re

Quotation marks used to doubt: the groups' independence is put in scare quotes in the headline; both groups' denials are quoted at length in the body — “Anthropic allies are behind ‘independent’ groups ringing alarm bells about AI”

Unnamed accusers: the central criticism is attributed to "critics" with no one named — “Critics have described Tarbell-backed journalists’ narratives about possible AI harms as exaggerated.”

Scare quotes in the headline: the groups' independence is marked as doubtful before any evidence is given — “Anthropic allies are behind ‘independent’ groups ringing alarm bells about AI”

Who is quoted: the two groups answer at length by name; the critics of them are unnamed — “Critics have described Tarbell-backed journalists’ narratives about possible AI harms as exaggerated.”

Right of reply: both accused organisations are quoted in their own words, one of them twice — ““We do not accept donations by frontier AI companies or their employees, and our funders have no say in the projects we work on,” METR President Chris Painter told the Washington Examiner.”

Read it at Washington Examiner → https://www.washingtonexaminer.com/news/investigations/4729617/anthropi…

Washington Examiner — independent

China won’t pause AI development, so the US would be insane to

Argues the US "would be insane" to pause, calls the people asking for one "performing for public opinion", and writes a chain of future Chinese military and economic actions in the plain indicative with nothing

The newest cultural-political fad? Call for a “pause” in artificial intelligence development. As columnist Dominic Pino observes, politicians can’t actually tell us what this “pause” should entail. That’s a tell that these pausers are performing for public opinion. On this issue at least, President Donald Trump sees the bigger picture. He knows that America currently holds the global initiative on AI. He knows that the United States must retain that initiative or suffer its gravest defeat in history. Recommended Stories China won’t pause AI development, so the US would be insane to Abdul el Sa

Contested claim stated flat: a genocide finding is asserted inside a rhetorical question, unattributed — “a nation which wages a genocide against its own people”

Speculation written in the indicative: a hypothetical future is given in "it would mean", with no source and no hedge — “It would mean the People’s Liberation Army being able to blind U.S. military communications and penetrate American defenses with hyperintelligent missile and drone swarm attacks.”

Contested characterisation stated flat: a genocide finding is asserted in the writer's own voice, unattributed, as a rhetorical question — “Does anyone seriously expect that a nation which wages a genocide against its own people would be a beneficent global master of the AI industry?”

Read it at Washington Examiner → https://www.washingtonexaminer.com/opinion/beltway-confidential/4729790…

The New York Times — independent

Why Is It So Difficult To Regulate A.I.?

The one report that treats the stalling itself as the question, and has a reporter explain it rather than a columnist.

There has been considerable debate over regulating the artificial intelligence industry, but much of that effort has stalled. Cecilia Kang, a technology reporter for The New York Times, explains why.

Read it at The New York Times → https://www.nytimes.com/video/business/100000011155275/why-is-it-so-dif…

National Review — independent

China Wants America to Fear AI. Congress Shouldn’t Encourage That

Tells Congress what to do rather than what it did: "distinguish between a guardrail and a roadblock — and understand who benefits from each."

Congress needs the discipline to distinguish between a guardrail and a roadblock — and understand who benefits from each.

Instruction to lawmakers: the standfirst tells the legislature what it needs rather than what it did — “Congress needs the discipline to distinguish between a guardrail and a roadblock — and understand who benefits from each.”

Read it at National Review → https://www.nationalreview.com/2026/09/china-wants-america-to-fear-ai-c…

T The Washington Post — independent

As warnings mount, Democrats eyeing the presidency wake up to AI fears

The only report on what Democrats eyeing the presidency are actually proposing: plans "heavier on rhetoric than specifics".

Their plans are heavier on rhetoric than specifics so far. But it’s a shift for many in the party who have been slow to address the fast-changing technology.

Read it at The Washington Post → https://www.washingtonpost.com/politics/2026/09/17/amid-warnings-human-…

T The Washington Post — independent

Trump won by latching on to popular issues. He’s doing the opposite on AI.

The only report claiming a public-opinion verdict — Trump on "the unpopular side" of the AI and data-centre debate — and it gives no poll figure or method.

It's been stunning to see this president so firmly position himself on the unpopular side of the debate around AI and data centers.

Popularity claimed without a number: which side is "unpopular" is asserted; no poll, method or figure is given — “It's been stunning to see this president so firmly position himself on the unpopular side of the debate around AI and data centers.”

Judgement in the lede: the president's position is called unpopular with no polling figure or method attached — “It's been stunning to see this president so firmly position himself on the unpopular side of the debate around AI and data centers.”

Read it at The Washington Post → https://www.washingtonpost.com/politics/2026/09/17/trump-won-by-latchin…

Checked, didn’t cover

Mother Jones

Reuters N NPR P Politico

Fox News

Breitbart

MSNBC

The Hill

Silence is recorded, not inferred — each of these was read when we built this story and carried nothing on this event.

We could not read these

The Wall Street Journal

Associated Press

These are must-cite outlets for this desk that we did not manage to read when we built this story. That is a gap in our reach, not a silence on their part, and it is not evidence of anything about this story.

Read another desk

Sections

Registers

Desks

Washington Examiner puts ‘independent’ in quote marks; METR says ‘our funders have no say in the projects we work on’

What is being argued about

Who is quoted about AI risk

Stated flat, or hedged

Claims and verdicts

Who wrote it

Coverage, outlet by outlet