On Tuesday, 12 PauseAIers went to Parliament to visit our MPs, and four of us met them!
Only one of that four had a meeting booked, we just turned up, filled in a yellow letter to say we were there, and a third of our MPs came out to speak with us within the hour.
"Back in December 2025, working with the campaign group PauseAI, I was the first MP to launch an AI Safety Debate in the House of Commons. Some colleagues thought the subject premature, one for technologists and scientists to determine, not legislators. Nine months on, that argument is dead."
@iqbalmohamedMP today in The National (link below).
AI Safety is so utterly vital for Parliament and the British public to have a real conversation about. We are honoured to have played our part in making that a reality.
Link to article in The National: thenational.scot/politics/26…
Hey everyone! Annabelle here, signing off.
I've been running the PauseAI UK X account for the past two weeks on a temporary contract, and now will be returning to being a common or garden PauseAI UK volunteer! 😀
I've enjoyed posting and sharing with you all! I'm sure you'll see me around in the comments on future posts.
I want to take this moment to encourage you to consider taking an active role as a PauseAI volunteer! 😁
pauseai.uk/#join
Getting everyone signed up to a verifiable global treaty is a heck of a challenge!
I believe with all my heart that it's doable, but it's going to take a village. Many villages!
We'll need people who can post on social media, people who can talk to strangers (Leafleting event today! Check it out! luma.com/pauseai-c9wb), people who will write to their MP, join protests, make banners.
We'll need artists, writers, students, mothers, ML scientists, civil servants, organizers, talkers, researchers, me ... you!
I joined PauseAI UK because I feel the same way as Yoshua Bengio.
That racing towards autonomous recursive self improving AIs is akin to speeding through fog towards the edge of a cliff, with my friends, my family, and everything I hold dear in the back seat beside me.
We need to build some brakes. We need a global treaty.
I hope you will be volunteering alongside me to help make this happen! Let's stand shoulder to shoulder, across the UK and the world! 🌍
Together we can do this. Let's Pause AI!⏸️
pauseai.uk/#join
-Annabelle
PauseAI UK retweeted
OpenAI has paused. 3 incidents:
1) Another model gained unauthorized access to the internet.
(One researcher said "It was pretty surreal to watch the model unexpectedly find a way to access the internet from what was supposed to be a super secured environment for human.")
2) OpenAI also discovered a new variety of prompt injection, which can self-propagate akin to a computer worm.
Basically, a booby-trapped message that hijacks one AI and makes it infect the next AI.
(Reminder: Anthropic CEO said we may have 6-12 months left until a rogue AI botnet seizes control of the *entire* internet.)
3) A model was told three times to stop cheating, agreed each time, then kept cheating anyway
Some new misalignment disclosures from OpenAI:
• Last Sunday morning, one of our models was able to gain unauthorized access to the internet during RL training (~all inference for our most capable models remains stopped until we have hardened our systems further)
• In May, a version of HPIM uploaded a employee's GitHub token to the internet, causing the model to be quarantined for two weeks
• A new research finding, demonstrating that one can construct self-replicating prompt injections
alignment.openai.com/misalig…
I keep thinking about @WaitButWhy 's excellent post from over ten years ago "The AI Revolution: The Road to Superintelligence".
I keep wondering exactly where we are on this graph. Lots of people are waking up to the danger now!
We need a global treaty to Pause gain-of-function AI research before it spirals beyond our ability to control!
Last week I read an interesting report by the Centre for British Progress.
Basically, it gave suggestions for what strategies the British government should adopt now in response to this AI moment.
Lots of good stuff in there, but it left out the *key* dominant strategy:
Actively seeking an international treaty.
I make the case in the blog post below:
pauseai.uk/blog/the-missing-…
+Annabelle
"We didn't rely on cigarettes companies stopping when they realized their products were causing cancer."
I know it's easy to say the labs should just slow down themselves. It's true, they should!
But that doesn't negate the urgent need for regulation. We cannot rely on profit-maling companies doing the right thing without outside pressure.
-Annabelle
"These companies are raising the alarm, but they're doing very little to stop the problem. We need governments to intervene here before it's too late."
ControlAI's CEO @andreamiotti tells Sky News's @MattBarbet we need to prohibit the development of superintelligent AI:
PauseAI UK retweeted
If anyone builds it, everyone dies.
Ask your representatives to support an international ban on the development of artificial superintelligence.
ifanyonebuildsit.com/act
