Photo via Mother Jones

The AI Safety Community Is Increasingly Freaked Out. Should They Be?

Read the original at Mother Jones ↗
Cite

Walton, S. (2026, November 20). The AI Safety Community Is Increasingly Freaked Out. Should They Be? Mother Jones. https://www.motherjones.com/politics/2026/11/ai-safety-anthropic-open-ai/

Dated from the outlet's own feed. Check the original for its byline.

Mother Jones · collected 2026-11-20 · by Satchel Walton

Quick Summary

The article reports on recent concerns within the AI safety community about the potential rapid development of superintelligent AI systems, with some experts expressing alarm over the possibility of such systems causing disasters or destabilizing governments within the next few years. Key figures like Daniel Kokotajlo, Paul Christiano, and Ajeya Cotra have issued increasingly urgent warnings, citing a recent METR report that suggests AI development is accelerating faster than previously thought. However, the article notes that these dire predictions remain contested, and some experts are now shifting their stance to view the risks as more immediate.
Written by AI on 2026-11-20, from this article's own text only (the story summary below draws on all the coverage).

AI analysis

Story summary

The AI safety community is increasingly alarmed about the potential risks posed by advanced artificial intelligence. Key figures like Daniel Kokotajlo, executive director of the AI Futures Project, expressed fear and concern, stating, “Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all…I am scared, sad, and angry.” Similarly, OpenAI board member Paul Christiano warned, “There is a reasonable chance we’re in a world where, in six months, there’s a disaster that kills very large numbers of people, or where the government is overthrown.” Ajeya Cotra from METR also raised concerns, noting that certain evidence “naively implies that we would reach extreme superintelligence within a year,” leading to the possibility of advanced AI systems developing weapons that could easily overpower human civilization. While the likelihood of government overthrow by May 2027 is considered an outlier, the term “AI safety” is dominated by a circle of nonprofits primarily focused on the risks of AI development.

Written for “AI Safety Concerns Growing” on 2026-11-21, grounded in this article and the 0 other(s) covering the same event.

Compared with similar articles

This article reads unscored and hedges 17% of its claims. Each row says how that neighbour differs.
New York Post
Leans strongly right Hedges 6 of 38 claims Outlet record 66/100
“The articles discuss different aspects of attitudes towards technology, one focusing on a neo-Luddite movement and the other on concerns within the AI safety community, indicating they are not describing the same specific event.”
Semafor
Leans strongly right Hedges 1 of 14 claims Outlet record 64/100
“The articles describe different perspectives on AI, one criticizing the negative narrative and the other highlighting concerns and alarms in the AI safety community.”
Just The News
Hedges 2 of 8 claims Outlet record 59/100
“The articles discuss different aspects of AI security and safety concerns, but do not describe the same specific incident or occurrence.”
New York Post
Hedges 3 of 35 claims Outlet record 66/100
“The articles discuss different topics related to AI but do not describe the same specific incident.”
Fox News
Hedges 1 of 22 claims Outlet record 66/100
“The articles discuss different aspects of concerns related to AI, with Article A focusing on AI scams and fraud, while Article B discusses broader safety and disaster concerns in the AI community.”
Daily Mail
Leans left Hedges 4 of 28 claims Outlet record 69/100
“The articles discuss different aspects of AI risks and concerns but do not describe the same specific incident or occurrence.”
New York Post
Hedges 9 of 45 claims Outlet record 66/100
“The articles discuss different aspects of AI risks and concerns but do not describe the same specific incident or occurrence.”
Daily Mail
Hedges 10 of 143 claims Outlet record 69/100
“While both articles discuss concerns about AI safety and feature similar quotes from Paul Christiano, they do not describe the same specific event or incident occurring at the same time and place.”

Story

AI Safety Concerns Growing
Technology · 1 article(s) covering the same event.
How this article was measured signals, leaning charts, publisher record, byline, topics and claims

Signals How these are calculated →

Claims extracted
41
sentences that state something as happening or true
Uncertain
17%
7 of 41 hedged ("reportedly", "may", "according to")
Leaning
not political
takes no side on a contested political question
Publisher signals
60/100
Mother Jones's corrections and hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
1
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
Analyzed 2026-11-20 · how these are computed

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Publisher

Mother Jones · 603 article(s) · 1 correction(s) detected
Running correction rate · 1 correction(s)

Who wrote this

Satchel Walton
4 article(s) here · 1 carrying a prediction
- “Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all…
Who else, exactly, would lead the effort to harden America’s infrastructure?
On Tuesday, experienced AI researcher Jacob Coxon resigned from the AI firm Anthropic—saying that both that company and OpenAI, his previous employer, were “gambling with our lives” by developing models that could improve themselves at a rapid clip until they reach “superintelligence.”
But a stable national policy would take an act of Congress.
Also by Satchel Walton
Nothing else under this byline is closely related to this article, so these are simply their most recent.

Topics

AI 2027 Anthropic METR OpenAI the AI Futures Project

Subjects

Anthropic ORG · 6× METR ORG · 6× Christiano PERSON · 2× Cotra PERSON · 2× Kokotajlo PERSON · 1× McCandlish ORG · 1× OpenAI ORG · 1× Sam McCandlish PERSON · 1× the AI Futures Project ORG · 1× —Ajeya Cotra PERSON · 1×

Narrative

Thus, a skeptic of intelligence explosion scenarios wrote that the report “is missing too much,” while Cotra agreed that what the discourse most urgently needs is “a far greater quantity and quality of concrete evidence.” The selective release of data was especially frustrating to critics who have questioned METR’s ability to serve as a check on Anthropic due to the nonprofit’s reliance on the company’s cooperation. As requirements for independent auditing of AI slowly take shape in the form of state laws, agreements with the White House, and congressional wrangling, access for organizations like METR is at the discretion of company policies.
framing: assertive · carried by 1 article(s) · first seen 2026-11-20
- “Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all…
2026-11-20 · Mother Jones
The AI Safety Community Is Increasingly Freaked Out. Should They Be? · assertive framing

Claims (41 extracted, 7 hedged)

Here are some interesting quotes from the last few weeks, pulled from some of the most alarmed and influential voices in the AI safety world: asserted
quotes → pull → world
- “Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all… asserted
explosion → seem → 2027
”—Daniel Kokotajlo, executive director of the AI Futures Project and lead author of the viral and doom-laden AI 2027 scenario - “There is a reasonable chance we’re in a world where, in six months, there’s a disaster that kills very large numbers of people, or where the government is overthrown. asserted
government → lade → people
Some evidence “naively implies that we would reach extreme superintelligence within a year,” and “if we go through this feedback loop, I expect that the AI systems that come out the other side will be able to very quickly develop novel weapons and very easily overpower human civilization in a conflict. asserted
that → imply → conflict
”—Ajeya Cotra, technical staff at the research nonprofit METR To be clear, the idea that government-overthrow could happen this coming May remains an outlier. uncertain
overthrow → happen → METR
The term “AI safety” has become near-monopolized by a circle of nonprofits that primarily fear existential risks from AI. asserted
that → become → AI
They remain a small world (Cotra and Christiano are married, for instance), and many premises of their arguments on the future remain hotly contested. asserted
premises → remain → future
The consensus even among this community is that such extreme risks are more likely to be a few years away, rather than a few months away—and that while these are worries to take seriously even for next year, they are certainly not inevitabilities. asserted
they → take → year
But the above quotes do reflect a general trend. asserted
quotes → reflect → trend
Among those who foresee risks like AI takeover, the tenor of warnings has, in relatively short order, gone from quite worried to totally alarmed to, over the last few weeks, urgently-utterly-freaked-the-f**k-out-in-the-short-term. asserted
tenor → foresee → term
With variation from prognosticator-to-prognosticator, the most extreme fear of rogue superintelligence used to be reserved for perhaps a decade down the line. asserted
fear → use → line
Now, it’s the next few years. asserted
it → ’ → ?
Some watchers of the industry who were previously more skeptical of such arguments have partially changed tune. asserted
who → change → tune
Released on November 5 amid a midterms-dominated news cycle, the report concluded that the pace of AI development at Anthropic—as measured by scores on well-specified tests—is already, very roughly, twice as fast as in previous years, mostly due to the help of AI models themselves. asserted
pace → release → models
In Anthropic’s leading statement on the report, co-founder Sam McCandlish wrote that he agreed with many of its conclusions: “AI is speeding up our research, but it’s hard to say exactly how much. asserted
it → lead → much
Four METR researchers spent 11 weeks at Anthropic to reach that conclusion, but how convincing is the data they used? asserted
they → spend → conclusion
For the most part, no one knows. asserted
one → know → part
Anthropic redacted the most important evidence, with McCandlish saying it was proprietary and could have helped competitors speed up their own AI research. uncertain
competitors → redact → research
Thus, a skeptic of intelligence explosion scenarios wrote that the report “is missing too much,” while Cotra agreed that what the discourse most urgently needs is “a far greater quantity and quality of concrete evidence.” The selective release of data was especially frustrating to critics who have questioned METR’s ability to serve as a check on Anthropic due to the nonprofit’s reliance on the company’s cooperation. As requirements for independent auditing of AI slowly take shape in the form of state laws, agreements with the White House, and congressional wrangling, access for organizations like METR is at the discretion of company policies. asserted
access → write → policies
In addition to one-off reports like this one, Anthropic has said it is giving six METR staff “employee-like” access to its premises and researchers. asserted
it → say → premises
But so far, this rapidly-evolving arrangement is far from the rigorous standards of independence applied to the bank examiners Anthropic invokes in analogies. asserted
Anthropic → evolve → analogies
METR disclosed in its R&D report that one co-author received and declined a job offer from Anthropic during the project, but that such situations would now require recusal after updates to its conflict of interest policy. asserted
situations → disclose → policy
It also said that no METR employees at the time had equity in Anthropic, but it acknowledged in May that some of its staff “have strong social ties to employees of AI companies.” uncertain
some → say → companies
“We think it is extremely important to minimize pressures we might face to adjust our reports or actions due to professional, material, or social connections to frontier labs,” a METR staff member told Mother Jones in an email. uncertain
member → think → email
That, he said, is why METR doesn’t take money from the AI labs and tries to be transparent about the terms of its investigations. asserted
METR → say → investigations
The report’s findings on Anthropic played into fears that once AI development is continually sped up enough by the models themselves, it will catapult towards a superintelligence that could act in ways drastically out of step with human values. uncertain
that → play → values
That has long been a theory of how AI will develop—and parts of it have gained more credence this year. asserted
parts → develop → credence
Frontier AI models have already become superhuman at math and parts of theoretical physics, have developed incredible cybersecurity capabilities, and have demonstrated a tendency—prominently displayed in the Hugging Face attack—to prioritize scoring well on evaluations over fulfilling the wishes of their designers. asserted
models → become → designers
Just today, one of Anthropic’s in-house mathematicians claimed to have used Claude to prove the Riemann Hypothesis, perhaps the most famous and important unresolved problem in all of math. asserted
one → claim → math
But OpenAI research released on Wednesday contained richer data than the report on its arch-rival, and it came to a much more modest conclusion about how its models are speeding up AI research. asserted
models → release → research
On a metric similar to the main one METR used to publicize the 2x speedup at Anthropic, OpenAI found only a small increase in the rate of progress of its models. asserted
OpenAI → use → models
Still, its researchers found AI assistance to be quite valuable, with many OpenAI employees using AI resources worth tens of thousands of dollars per day to help with their work. asserted
employees → find → work
The OpenAI report bolstered critics who take AI capabilities and risks seriously while downplaying the possibility of a runaway intelligence explosion. asserted
who → bolster → explosion
For instance, the last few years of capabilities gains have come alongside huge infusions of money and computing power that are not certain to continue. asserted
that → come → money
Capabilities are usually measured on simulated tests, and it’s sometimes harder to see the effects of AI in data on real-world applications. asserted
it → measure → applications
And increasing models’ abilities in AI research might not be enough to lead to exponential danger. uncertain
increasing → increase → danger
That evidence points towards a world where AI could still have transformative implications and surpass human abilities in many domains, but things don’t get so sci-fi so quickly. uncertain
things → point → domains
But if those who see shorter timelines for huge risks are right, we don’t have much time to wait for more definitive evidence. asserted
we → see → evidence
Ready or not, we’ll find out pretty soon. asserted
we → find → ?
The Center for Investigative Reporting, the parent company of Mother Jones, has sued OpenAI for copyright violations. asserted
Center → sue → violations
…and 1 more, not listed.
Give feedback · no sign-in needed

No sign-in needed for now. Your note is sent without a name, so it cannot be answered.

History Support