ControlAI
@controlai.com
We work to keep humanity in control.
Subscribe to our free newsletter: https://controlai.news
Join our discord at: https://discord.com/invite/ptPScqtdc5
Subscribe to our free newsletter: https://controlai.news
Join our discord at: https://discord.com/invite/ptPScqtdc5
Pinned
ControlAI
@controlai.com
· May 29
I’m sorry Sam, I’m afraid I can’t do that
Blackmail and sabotage
controlai.news
AIs are sabotaging their own shutdown mechanisms, blackmailing engineers, and showing concerning biothreat capabilities.
Meanwhile, an Anthropic researcher admits they want Claude n to build Claude n+1.
Covered in the latest edition of our newsletter:
controlai.news/p/im-...
Meanwhile, an Anthropic researcher admits they want Claude n to build Claude n+1.
Covered in the latest edition of our newsletter:
controlai.news/p/im-...
While many of you have been writing your Christmas cards, all the top AI companies have published different kinds of cards — system cards.
These documents, which outline the capabilities, limitations, and safety measures of an AI system, paint an interesting picture.
Thread 🧵
These documents, which outline the capabilities, limitations, and safety measures of an AI system, paint an interesting picture.
Thread 🧵
November 30, 2025 at 5:09 PM
While many of you have been writing your Christmas cards, all the top AI companies have published different kinds of cards — system cards.
These documents, which outline the capabilities, limitations, and safety measures of an AI system, paint an interesting picture.
Thread 🧵
These documents, which outline the capabilities, limitations, and safety measures of an AI system, paint an interesting picture.
Thread 🧵
Max Tegmark says we have no clue how to control artificial superintelligence.
If we can prevent that from being developed, we can build a good future.
If we can prevent that from being developed, we can build a good future.
November 30, 2025 at 10:35 AM
Max Tegmark says we have no clue how to control artificial superintelligence.
If we can prevent that from being developed, we can build a good future.
If we can prevent that from being developed, we can build a good future.
MIRI CEO Malo Bourgon explains why AIs would try to preserve themselves, acquire resources, and resist their goals being changed.
"This was kind of theoretical 10 years ago."
"Now we're starting to see that ... these behaviors are starting to manifest."
"This was kind of theoretical 10 years ago."
"Now we're starting to see that ... these behaviors are starting to manifest."
November 29, 2025 at 5:47 PM
MIRI CEO Malo Bourgon explains why AIs would try to preserve themselves, acquire resources, and resist their goals being changed.
"This was kind of theoretical 10 years ago."
"Now we're starting to see that ... these behaviors are starting to manifest."
"This was kind of theoretical 10 years ago."
"Now we're starting to see that ... these behaviors are starting to manifest."
MIRI CEO Malo Bourgon explains why AI isn't like other technologies, and why it looks likely that superintelligence will be developed much earlier than previously thought:
November 29, 2025 at 10:54 AM
MIRI CEO Malo Bourgon explains why AI isn't like other technologies, and why it looks likely that superintelligence will be developed much earlier than previously thought:
We have a new supporter!
Stewart Dickson MLA just backed our campaign calling for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
98 UK politicians now support our campaign!
Stewart Dickson MLA just backed our campaign calling for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
98 UK politicians now support our campaign!
November 28, 2025 at 3:53 PM
We have a new supporter!
Stewart Dickson MLA just backed our campaign calling for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
98 UK politicians now support our campaign!
Stewart Dickson MLA just backed our campaign calling for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
98 UK politicians now support our campaign!
Not the Christmas cards you were hoping for.
New system cards from OpenAI, Anthropic, Google and xAI indicate AIs are becoming more capable in dangerous domains such as biological weapons and automating AI research.
Read more in our latest article!
New system cards from OpenAI, Anthropic, Google and xAI indicate AIs are becoming more capable in dangerous domains such as biological weapons and automating AI research.
Read more in our latest article!
Dangerous AI Capabilities Advance
“My trust in reality is fading” — Gemini
controlai.news
November 27, 2025 at 7:02 PM
Not the Christmas cards you were hoping for.
New system cards from OpenAI, Anthropic, Google and xAI indicate AIs are becoming more capable in dangerous domains such as biological weapons and automating AI research.
Read more in our latest article!
New system cards from OpenAI, Anthropic, Google and xAI indicate AIs are becoming more capable in dangerous domains such as biological weapons and automating AI research.
Read more in our latest article!
We now have 97 supporters!
Bob Blackman MP just backed our campaign for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
It's great to see this coalition keep growing!
Bob Blackman MP just backed our campaign for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
It's great to see this coalition keep growing!
November 26, 2025 at 5:20 PM
We now have 97 supporters!
Bob Blackman MP just backed our campaign for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
It's great to see this coalition keep growing!
Bob Blackman MP just backed our campaign for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
It's great to see this coalition keep growing!
In Anthropic’s system card for their newest AI, Claude Opus 4.5, they say that confidently ruling out that their AI R&D-4 dangerous capability threshold has been crossed is “becoming increasingly difficult”.
A thread on what this capability would mean 🧵
A thread on what this capability would mean 🧵
November 25, 2025 at 5:10 PM
In Anthropic’s system card for their newest AI, Claude Opus 4.5, they say that confidently ruling out that their AI R&D-4 dangerous capability threshold has been crossed is “becoming increasingly difficult”.
A thread on what this capability would mean 🧵
A thread on what this capability would mean 🧵
How easy is it to write to your representatives about the danger of superintelligence and the need to address it?
We’ve made it really easy! With our contact tools, it takes under a minute to complete the whole process.
Thousands have already used them!
Try them out below!
We’ve made it really easy! With our contact tools, it takes under a minute to complete the whole process.
Thousands have already used them!
Try them out below!
November 24, 2025 at 5:50 PM
How easy is it to write to your representatives about the danger of superintelligence and the need to address it?
We’ve made it really easy! With our contact tools, it takes under a minute to complete the whole process.
Thousands have already used them!
Try them out below!
We’ve made it really easy! With our contact tools, it takes under a minute to complete the whole process.
Thousands have already used them!
Try them out below!
AI researcher Nate Soares says even those building modern AI systems don't know what's going on inside them.
"They'll just come out and tell you."
"They'll just come out and tell you."
November 24, 2025 at 4:36 PM
AI researcher Nate Soares says even those building modern AI systems don't know what's going on inside them.
"They'll just come out and tell you."
"They'll just come out and tell you."
AI researcher Nate Soares says that despite AIs getting more powerful, we're not on track to be able to control them.
"This is sort of a worst case situation. I never wanted to be here."
"This is sort of a worst case situation. I never wanted to be here."
November 23, 2025 at 5:03 PM
AI researcher Nate Soares says that despite AIs getting more powerful, we're not on track to be able to control them.
"This is sort of a worst case situation. I never wanted to be here."
"This is sort of a worst case situation. I never wanted to be here."
Microsoft AI CEO Mustafa Suleyman says he thinks it's "quite likely" that AIs will surpass humans at doing most tasks within the next 5 years.
November 23, 2025 at 10:37 AM
Microsoft AI CEO Mustafa Suleyman says he thinks it's "quite likely" that AIs will surpass humans at doing most tasks within the next 5 years.
We just got our 96th supporter!
Lord Bethell just backed our campaign calling for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
It's great to see so many politicians take a stand on this issue!
Lord Bethell just backed our campaign calling for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
It's great to see so many politicians take a stand on this issue!
November 22, 2025 at 10:39 AM
We just got our 96th supporter!
Lord Bethell just backed our campaign calling for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
It's great to see so many politicians take a stand on this issue!
Lord Bethell just backed our campaign calling for binding regulation on the most powerful AI systems, acknowledging the extinction risk posed by AI.
It's great to see so many politicians take a stand on this issue!
John Whitby MP just backed our campaign for binding regulation on the most powerful AI systems!
95 UK politicians now support our campaign, acknowledging the extinction risk posed by AI.
It's great to see so many politicians coming together on this issue!
95 UK politicians now support our campaign, acknowledging the extinction risk posed by AI.
It's great to see so many politicians coming together on this issue!
November 21, 2025 at 3:01 PM
John Whitby MP just backed our campaign for binding regulation on the most powerful AI systems!
95 UK politicians now support our campaign, acknowledging the extinction risk posed by AI.
It's great to see so many politicians coming together on this issue!
95 UK politicians now support our campaign, acknowledging the extinction risk posed by AI.
It's great to see so many politicians coming together on this issue!
AI just helped run a highly sophisticated real-world cyber-espionage campaign at scale, across government and industry.
In our latest article, we break down what happened and what it means.
In our latest article, we break down what happened and what it means.
When the Hacker is AI
AI just helped run a highly sophisticated real-world cyber-espionage campaign at scale.
controlai.news
November 20, 2025 at 7:10 PM
AI just helped run a highly sophisticated real-world cyber-espionage campaign at scale, across government and industry.
In our latest article, we break down what happened and what it means.
In our latest article, we break down what happened and what it means.
Baroness Boycott just backed our UK campaign!
94 UK politicians now support binding regulation on the most powerful AI systems, acknowledging that AI poses an extinction risk to humanity.
Link below for our campaign statement and full list of supporters!
👇
94 UK politicians now support binding regulation on the most powerful AI systems, acknowledging that AI poses an extinction risk to humanity.
Link below for our campaign statement and full list of supporters!
👇
November 20, 2025 at 11:02 AM
Baroness Boycott just backed our UK campaign!
94 UK politicians now support binding regulation on the most powerful AI systems, acknowledging that AI poses an extinction risk to humanity.
Link below for our campaign statement and full list of supporters!
👇
94 UK politicians now support binding regulation on the most powerful AI systems, acknowledging that AI poses an extinction risk to humanity.
Link below for our campaign statement and full list of supporters!
👇
Microsoft AI CEO Mustafa Suleyman says it would be very hard to contain or align smarter-than-human AI that could self-improve, act independently, and set its own goals.
"That should be the anti-goal."
"That should be the anti-goal."
November 19, 2025 at 5:21 PM
Microsoft AI CEO Mustafa Suleyman says it would be very hard to contain or align smarter-than-human AI that could self-improve, act independently, and set its own goals.
"That should be the anti-goal."
"That should be the anti-goal."
We just got a new supporter!
Sir Desmond Swayne MP just backed our campaign to regulate the most powerful AI systems.
93 UK politicians now support our campaign statement, acknowledging the extinction threat posed by AI!
It's great to see this coalition grow!
Sir Desmond Swayne MP just backed our campaign to regulate the most powerful AI systems.
93 UK politicians now support our campaign statement, acknowledging the extinction threat posed by AI!
It's great to see this coalition grow!
November 18, 2025 at 11:20 AM
We just got a new supporter!
Sir Desmond Swayne MP just backed our campaign to regulate the most powerful AI systems.
93 UK politicians now support our campaign statement, acknowledging the extinction threat posed by AI!
It's great to see this coalition grow!
Sir Desmond Swayne MP just backed our campaign to regulate the most powerful AI systems.
93 UK politicians now support our campaign statement, acknowledging the extinction threat posed by AI!
It's great to see this coalition grow!
There's a great new article in BBC Arabic about the huge coalition of experts and leaders calling for a ban on superintelligence.
Why does it need to be prohibited?
Why does it need to be prohibited?
November 17, 2025 at 4:39 PM
There's a great new article in BBC Arabic about the huge coalition of experts and leaders calling for a ban on superintelligence.
Why does it need to be prohibited?
Why does it need to be prohibited?
Discussing the use of Anthropic's Claude AI to conduct cyberattacks, CEO Dario Amodei says that as well as being misused in this way, AI is going to "go wrong on its own".
November 17, 2025 at 12:21 PM
Discussing the use of Anthropic's Claude AI to conduct cyberattacks, CEO Dario Amodei says that as well as being misused in this way, AI is going to "go wrong on its own".
Microsoft AI CEO Mustafa Suleyman says there's a creeping assumption that it's inevitable that AI will exceed our control and move beyond us as a species.
November 16, 2025 at 5:59 PM
Microsoft AI CEO Mustafa Suleyman says there's a creeping assumption that it's inevitable that AI will exceed our control and move beyond us as a species.
Elon Musk says he doesn't think anyone will have control over superintellignece any more than a chimp would have control over humans.
November 16, 2025 at 10:54 AM
Elon Musk says he doesn't think anyone will have control over superintellignece any more than a chimp would have control over humans.
AI godfather Geoffrey Hinton explains why he decided to start warning that AI could wipe us out.
At the time, people were saying AIs are stochastic parrots.
Hinton says that's wrong. "They really do understand what you're saying. And they really are gonna get smarter than us."
At the time, people were saying AIs are stochastic parrots.
Hinton says that's wrong. "They really do understand what you're saying. And they really are gonna get smarter than us."
November 15, 2025 at 4:49 PM
AI godfather Geoffrey Hinton explains why he decided to start warning that AI could wipe us out.
At the time, people were saying AIs are stochastic parrots.
Hinton says that's wrong. "They really do understand what you're saying. And they really are gonna get smarter than us."
At the time, people were saying AIs are stochastic parrots.
Hinton says that's wrong. "They really do understand what you're saying. And they really are gonna get smarter than us."
We have a new supporter!
Lord Randall is the 92nd UK politician to back our campaign for binding regulation on the most powerful AI systems, acknowledging that AI poses a risk of human extinction.
We're delighted that so many more are joining!
Lord Randall is the 92nd UK politician to back our campaign for binding regulation on the most powerful AI systems, acknowledging that AI poses a risk of human extinction.
We're delighted that so many more are joining!
November 15, 2025 at 10:23 AM
We have a new supporter!
Lord Randall is the 92nd UK politician to back our campaign for binding regulation on the most powerful AI systems, acknowledging that AI poses a risk of human extinction.
We're delighted that so many more are joining!
Lord Randall is the 92nd UK politician to back our campaign for binding regulation on the most powerful AI systems, acknowledging that AI poses a risk of human extinction.
We're delighted that so many more are joining!
AI godfather Geoffrey Hinton explains why he signed the call to ban superintelligence, which now has over 120,000 supporters.
"If we know we can't do it safely, we should stop.
And maybe if that knowledge is widely percolated to the public, we will be able to stop."
"If we know we can't do it safely, we should stop.
And maybe if that knowledge is widely percolated to the public, we will be able to stop."
November 14, 2025 at 2:35 PM
AI godfather Geoffrey Hinton explains why he signed the call to ban superintelligence, which now has over 120,000 supporters.
"If we know we can't do it safely, we should stop.
And maybe if that knowledge is widely percolated to the public, we will be able to stop."
"If we know we can't do it safely, we should stop.
And maybe if that knowledge is widely percolated to the public, we will be able to stop."