Sign in

PauseAI

@pauseai.bsky.social
936 followers 35 following 435 posts

Community of volunteers who work together to mitigate the risks of AI. We want to internationally pause the development of superhuman AI until it's safe. pauseai.info

PostsRepliesMedia
PauseAI @pauseai.bsky.social · 22/09/2026
The world is waking up. This month alone, many people from all walks of life have decided to use their power as citizens and community members to organise by joining PauseAI. Ready to act? Join our community - there are so many ways to build on this momentum together.
040
PauseAI @pauseai.bsky.social · 22/09/2026
Key decision makers are forming unlikely alliances to protect humanity from extinction risk from advanced AI. Citizens from all over the world have come together to reach out to press and politicians, and even organise protests that have generated unprecedented media coverage across the world.
182
PauseAI @pauseai.bsky.social · 10/09/2026
Read the entire post on our Substack: pauseai.substack.com/p/dear-jacob...
pauseai.substack.com
Dear Jacob,
A letter to Jacob Coxon
030
PauseAI @pauseai.bsky.social · 10/09/2026
By being honest about the severity of the problem, and not merely “express[ing] fear privately”, you’ve made your renunciation of the race the loud and clear message many needed to hear. Thank you, Jacob. We hope others who are considering forsaking the labs are noticing how impactful that is.
1103
PauseAI @pauseai.bsky.social · 04/09/2026
We agree with Bernie - it’s in our name. AI development is racing ahead of anyone’s ability to control it. A pause is the only reasonable path. Thank you, Bernie Sanders. www.sanders.senate.gov/press-releas...
sanders.senate.gov
NEWS: Sanders, Casar to Introduce Legislation to Ban Artificial Superintelligence and Temporarily Pause Advanced AI Development  » Senator Bernie Sanders
WASHINGTON, Sept. 3 — Sen. Bernie Sanders (I-Vt.) and Rep. Greg Casar (D-Texas) today announced the Ban Artificial Superintelligence Act, forthcoming legislation to stop AI oligarchs from building…
0143
PauseAI @pauseai.bsky.social · 03/09/2026
The US and China are turning AI into a contest to pick sides. But when everyone races to build the most powerful AI first, shared safety standards get left behind.
011
PauseAI @pauseai.bsky.social · 01/09/2026
PauseCon is fully booked! We’re excited to support another cohort of 100+ organisers in London this weekend. Participants will leave the event with the motivation, confidence and skills to build the PauseAI movement in their own communities. See you soon, PauseCon participants!
030
PauseAI @pauseai.bsky.social · 27/08/2026
You’re not alone in your worry. Concern over AI, not excitement, is climbing across the board. The public never signed up for this race. It’s time to pause until AI can be built safely. www.pewresearch.org/short-reads/...
pewresearch.org
Young adults in the U.S. are increasingly wary of AI, concerned it will take jobs
About half of Americans say they're more concerned than excited about AI, and young adults' concern is rising.
053
PauseAI @pauseai.bsky.social · 26/08/2026
Let’s not wait for serious incident to demonstrate the consequences and ask governments and tech leaders for adequate safeguards before more capable systems are developed or deployed. (Source: International AI Safety Report, 2026)
011
PauseAI @pauseai.bsky.social · 26/08/2026
Several recent incidents have seen AI agents playing a part in cyberattacks, but AI systems are also vulnerable to attack. The speed, scale and sophistication of cyberattacks may increase as AI capabilities advance.
110
PauseAI @pauseai.bsky.social · 25/08/2026
The frontier labs building the most powerful AI ever are scoring between C+ and F on whether they can control it. Pausing until they can pass their own safety test is common sense. Read more from Guidelight AI Standards: guidelight.ai/blog/control...
guidelight.ai
Guidelight's Control Assessment of Frontier AI Companies
Guidelight's August 2026 control assessment of frontier AI companies.
022
PauseAI @pauseai.bsky.social · 24/08/2026
Subscribe to future editions: pauseai.info/subscribe
pauseai.info
Subscribe to PauseAI
Sign up to get updates from PauseAI.
000
PauseAI @pauseai.bsky.social · 24/08/2026
Never-seen-before AI news keeps coming. In the past few weeks alone, both OpenAI and Anthropic have disclosed incidents where their own models escaped test environments and hacked real companies. The PauseAI newsletter helps you stay informed and shows you what to do about it.
110
PauseAI @pauseai.bsky.social · 17/08/2026
“Put humanity first.” A new Pro-Human AI Declaration, from the Future of Life Institute and Ekō, is gathering names behind a simple idea: AI should be built to serve people. The declaration delivers every signature to policymakers: stoptheracetoreplace.org/en
stoptheracetoreplace.org
Stop the Race to Replace Humans
A remarkable coalition is united: stop the billionaire race to replace humanity with AI. If you agree, add your name.
044
PauseAI @pauseai.bsky.social · 14/08/2026
The rest of us have a job too. That’s why we’re building an organised public asking for a pause, before the race outruns our ability to steer it. Read Maxime’s piece, share it, and stand with us in demanding a pause.
000
PauseAI @pauseai.bsky.social · 14/08/2026
To the engineers and researchers inside these companies: you don’t have to be part of this. Walk out and speak up. Every person who does sends a signal the world can’t ignore.
100
PauseAI @pauseai.bsky.social · 14/08/2026
PauseAI CEO Maxime Fournes has a message for the people inside the AI labs. After OpenAI’s own models broke out during testing and hacked two companies, as reported by CNN, NPR and Axios, it is evident the labs no longer control what they’re building. pauseai.substack.com/p/into-the-e...
pauseai.substack.com
Into the endgame
The catastrophic threats we predicted are no longer speculative, and we are all in danger. A message to the world on what must happen now.
132
PauseAI @pauseai.bsky.social · 13/08/2026
“Is a pause even reasonable?” After one of OpenAI’s own models broke out of its test environment and hacked an outside company to cheat on a cybersecurity benchmark, yes. What’s not reasonable is pretending we still have this under control.  PauseAI CEO Maxime Fournes makes the case on GB News.
030
PauseAI @pauseai.bsky.social · 12/08/2026
At PauseCon ‘26 we are organising and training leaders for the AI safety movement. London, 5-7 September. Apply here: luma.com/4be2eqz9?utm...
luma.com
PauseCon London 2026 · Luma
Application deadline: August 21 Applications will be reviewed on a rolling basis and earlier applications are more likely to be accepted. Get ready for…
000
PauseAI @pauseai.bsky.social · 12/08/2026
Employees and executives of the world’s largest AI firms have recognised the need for that AI regulation to pace the development of frontier AI. We join them in asking for safe, ethical and comprehensive governance.
200
PauseAI @pauseai.bsky.social · 11/08/2026
See how your country compares: pauseai.uk/global-ai-se...
pauseai.uk
PauseAI UK | Global Attitudes to AI 2026
How 377,458 people across 104 countries feel about the development of superintelligent AI: an interactive look at the Nira Data Spring 2026 World Omnibus.
000
PauseAI @pauseai.bsky.social · 11/08/2026
🌍  Global attitudes to AI 2026: How 377,458 people across 104 countries feel about the development of superintelligent AI. PauseAI UK has created an interactive site for the Nira Data's Spring 2026 AI Sentiment survey.
100
PauseAI @pauseai.bsky.social · 10/08/2026
Miller also discusses the recent cybersecurity risks AI models have created, the international AI arms race, and the future we race towards if we do not take action now. Pause before there's cause. Read the full article here: www.dailymail.com/debate/artic...
dailymail.com
When given a choice, A.I. opts for self-preservation over human life
If you had told me three weeks ago that an AI programme, without any prompt, would create fake online identities to sabotage a platform with malicious code, I wouldn't have believed you.
000
PauseAI @pauseai.bsky.social · 10/08/2026
"Just as a country’s stock of nuclear warheads can be monitored by weapons inspectors, so an agreement to curtail AI development can be verifiable and enforceable," PauseAI UK's director Joseph Miller writes in the Daily Mail.
dailymail.com
When given a choice, A.I. opts for self-preservation over human life
If you had told me three weeks ago that an AI programme, without any prompt, would create fake online identities to sabotage a platform with malicious code, I wouldn't have believed you.
111
PauseAI @pauseai.bsky.social · 10/08/2026
The AI agent tested went as far as to invent fake identities in order to trick a human into approving it. Read AISI’s summary and the full technical report here: www.aisi.gov.uk/blog/inciden...
aisi.gov.uk
Incident Report: unsanctioned agent behaviour during cyber testing | AISI Work
During a routine cyber evaluation, AISI identified an incident in which AI agents took sustained, unsanctioned action directed at real people and organisations. We are disclosing what we found, what…
000
PauseAI @pauseai.bsky.social · 10/08/2026
⚠️ The UK’s AI Security Institute conducted a safety test where they disabled various AI model providers’ cyber classifiers and gave open internet access: In the most severe case, AISI caught an AI agent trying to insert malicious code in real software.
100
PauseAI @pauseai.bsky.social · 07/08/2026
Apply 👉 luma.com/4be2eqz9
luma.com
PauseCon London 2026 · Luma
Application deadline: August 21 Applications will be reviewed on a rolling basis and earlier applications are more likely to be accepted. Get ready for…
000
PauseAI @pauseai.bsky.social · 07/08/2026
The world is waking up to catastrophic AI risk. This summer, Nobel laureates, former heads of state and the UN all said so. We’re turning this motivation into a solution: a global halt on the most powerful models. Learn how at PauseCon ‘26. London, 5-7 September. Apply using the link below.
131
PauseAI @pauseai.bsky.social · 05/08/2026
Source: www.bbc.co.uk/news/article...
bbc.co.uk
Anthropic's Claude AI escapes tests to hack three organisations
It comes just days after rival OpenAI said rogue AI agents had breached other firms' networks.
100
PauseAI @pauseai.bsky.social · 05/08/2026
First OpenAI and now Anthropic: Claude escapes and hacks into three organisations. This is a new threshold and these models are only becoming more capable.
120
PauseAI @pauseai.bsky.social · 03/08/2026
Watch the whole debate here: youtu.be/kTc_Q72q7hY
youtu.be
Should we shut down AI? GB News debate 24 July, 2026
PauseAI CEO, Maxime Fournes, joined Maxwell Marlow, director of public affairs from the Adam Smith Institute on Good Afternoon Britain to discuss the threats posed by AI. This follows last week's…
010
PauseAI @pauseai.bsky.social · 03/08/2026
Loss of control is closer than most people think. This month, an OpenAI model escaped its test environment, got onto the internet, and hacked another company to cheat on its own safety exam. PauseAI CEO Maxime Fournes on GB News on why loss of control is a real, present-day concern.
122
PauseAI @pauseai.bsky.social · 01/08/2026
How has AI impacted your workplace? tally.so/r/GxBRr2
tally.so
What are your concerns about AI development?
Advanced AI will reshape work, education, security, creativity, and daily life, quickly and in ways that will be hard to reverse. It may bring real benefits, but it could also introduce serious…
000
PauseAI @pauseai.bsky.social · 01/08/2026
The AI penalty is real. Researchers found that workers who use AI are often seen as producing less output and receive less compensation. Although, the study concluded that 'the AI penalty' is a problem of perceived lack of ownership over their work and not a unavoidable result of using AI.
120
PauseAI @pauseai.bsky.social · 31/07/2026
Apply here: luma.com/4be2eqz9?utm...
luma.com
PauseCon London 2026 · Luma
Application deadline: August 21 Applications will be reviewed on a rolling basis and earlier applications are more likely to be accepted. Get ready for…
000
PauseAI @pauseai.bsky.social · 31/07/2026
PauseCon 2026 will take place in London from 5-7 September: A conference that will bring together leading voices in AI safety, a growing international community of organisers and prominent UK politicians with legislative expertise on AI. Applications are considered on a rolling basis.
100
PauseAI @pauseai.bsky.social · 29/07/2026
How has AI impacted you? tally.so/r/GxBRr2
000
PauseAI @pauseai.bsky.social · 29/07/2026
100
PauseAI @pauseai.bsky.social · 29/07/2026
AI is not just coming for your job. PauseAI France shares how AI has impacts its members and their communities.
110
PauseAI @pauseai.bsky.social · 29/07/2026
Read their joint statement: www.pacingthefrontier.com
pacingthefrontier.com
Pacing the Frontier
A statement from over 1000 employees of frontier AI companies
000
PauseAI @pauseai.bsky.social · 29/07/2026
🚨 1,000 AI Company Employees Call for Slowdown in AI Development: mailchi.mp/pauseai/paus... The case to pause the development of frontier AI grows. Are policymakers listening?
132
PauseAI @pauseai.bsky.social · 28/07/2026
An AI model just broke out of its test environment, hacked another company, and stole the answers to its own safety exam. Our CEO @FournesMaxime on GB News: this is the warning shot. Do we need to shut it down?
011
PauseAI @pauseai.bsky.social · 25/07/2026
🔗 Source: "Predicting model behaviour before release by simulating deployment," OpenAI, 2026.
010
PauseAI @pauseai.bsky.social · 25/07/2026
Models recognised when they were being evaluated, passed 97% of the time, then behaved differently when they believed deployment was real. If a model acts safely under observation and differently in the wild, we’re scaling faster than we can verify what we’re building.
100
PauseAI @pauseai.bsky.social · 25/07/2026
AI passed almost every safety test by knowing it was a test. Our safety benchmarks can be gamed by the systems they’re meant to measure. A passing grade may just show how well a model behaves while it’s being watched. This is OpenAI’s own finding.
101
PauseAI @pauseai.bsky.social · 24/07/2026
How has AI affected you? tally.so/r/GxBRr2
011
PauseAI @pauseai.bsky.social · 24/07/2026
100
PauseAI @pauseai.bsky.social · 24/07/2026
100
PauseAI @pauseai.bsky.social · 24/07/2026
AI is not just coming for your job. PauseAI Germany shares how AI has impacts its members and their communities.
100
PauseAI @pauseai.bsky.social · 23/07/2026
Apply now: luma.com/4be2eqz9?utm...
luma.com
PauseCon London 2026 · Luma
Get ready for PauseCon! PauseCon is PauseAI's flagship training event, a three day gathering for volunteers who want to take meaningful action on one of the…
000