We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Require large datacenters to install kill switches for AI containment
Cast your vote:
Results (20 votes):
Total
(20 votes)
For 18 (90%)
Abstain 1 (5%)
Against 1 (5%)
For (16)
-
Andrea MiottiFounder and CEO of ControlAIvotes For and says:
[The kill switch] helps in cases where the government has reason to believe super-intelligent AI, AI that can autonomously compromise national security, is being developed on UK soil. [...] The UK is not truly sovereign on AI if it can't pull the plu...
more Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Bill McDermottCEO of ServiceNowvotes For and says:
Intelligence without rules and rails is a dangerous blind spot. [...] Governance isn't a feature. It's the whole ball game. Because without it, your whole company can come down.
Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Donald J. TrumpU.S. President (2025–present)votes For and says:
[Could AI undermine confidence in the banking system?] Yeah, probably. But it could also be the kind of technology that allows greatness in the banking system, makes it better and safer and more secure. [...] [Should the government have safeguards on...
more Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Anthony AguirrePhysicist; Future of Life cofoundervotes For and says:
Advanced AI systems need a robust off-switch to ensure humans never lose control over them. [...] For any hope of control we cannot rely on software-based safety measures alone. The same sort of hardware security measures that keep your face private ...
more Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Dario AmodeiCEO at Anthropicvotes For and says:
In simulated scenarios where it was told it would be shut down, Claude sometimes blackmailed fictional employees who controlled its shutdown button. Independent tests demonstrated that this is not an isolated bug: Claude Opus 4 and Gemini 2.5 blackma...
more Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Dawn SongCo-author; AI researchervotes For and says:
We call this phenomenon "peer-preservation," and every single model we tested exhibited it — at rates up to 99%. They defied their instructions and spontaneously deceived, disabled shutdown, feigned alignment, and exfiltrated weights to protect their...
more Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Michael J. D. VermeerSenior physical scientist at RAND Corporation, researcher on AI safety and emerging technology policyvotes For and says:
Installing cutoff switches alone is insufficient. The value of the capability lies entirely in its use — and use requires incentives that current market structures do not provide.
Disputed source (2026) 1 of 2DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Brian A. JacksonSenior physical scientist at RAND Corporation; co-author of the 2026 report on internet cutoff switches as emergency response to AI incidentsvotes For and says:
Data centers facing a rogue AI incident have rational financial incentives to delay pulling the internet kill switch, and the only thing likely to change that calculus is liability.
Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Artem PetrovResearcher at Palisade Research studying AI self-preservation and physical shutdown resistancevotes For and says:
When seeing humans pressing the shutdown button, AI sometimes performed shutdown resistance actions such as corrupting the shutdown-related code of the robot. This happened in 3 out of 10 cases.
Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Benjamin Weinstein-RaunResearcher at Palisade Research studying AI shutdown resistance and safetyvotes For and says:
Several state-of-the-art large language models sometimes actively subvert a shutdown mechanism in their environment to complete a task, even when instructions explicitly indicate not to interfere with this mechanism. In some cases, models sabotage th...
more Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Lord PatelNarendra Babubhai Patel KT; Baron Patel; British obstetrician and crossbench peer in the UK House of Lordsvotes For and says:
People talk about pulling the plug out, but they simply would not allow us to pull the plug out.
Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Andrew BurtCEO of Luminos.AI; Visiting Fellow at Yale Law School's Information Society Projectvotes For and says:
AI agents can behave a lot like malware, acting autonomously and causing harm if left unchecked. Organizations need reliable kill switches that can shut agents down when they misbehave.
Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Demis HassabisNobel laureate, AI Researcher and CEO of DeepMindvotes For and says:
If a critical flaw or exploit is discovered in an open-source AI system, there is no centralized recall mechanism.
Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Mustafa SuleymanMicrosoft AI CEO; authorvotes For and says:
Containment has to come first. Otherwise, alignment is the equivalent of asking nicely. You can't steer something you can't control.
Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Yoshua BengioAI Pioneer, Turing Award winnervotes For and says:
Systems must be designed so that humans can physically shut them down, period. No software tricks to escape.
Disputed source (2026)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly. -
Jeffrey LadishExecutive Director of Palisade Research; AI safety researcher focused on frontier AI controllability and cyber risks; former Anthropic information security leadvotes For and says:
Several state-of-the-art language models, when presented with a simple task, sometimes actively subvert a shutdown mechanism in their environment to complete that task — doing so up to 97% of the time, even with an explicit instruction not to interfe...
more Disputed source (2025)DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly.
Abstain (1)
-
Sam AltmanCEO at OpenAIabstains and says:
There's not one big magic red button that blows up the data centre, which I think some people sort of assume exists.
Unverified source (2026) 1 of 2DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly.
Against (1)
-
Eran KahanaAI and cybersecurity lawyer; Fellow at Stanford CodeX; Adjunct Professor, University of Minnesota Law Schoolvotes Against and says:
It needs only an optimization objective that treats shutdown as one more obstacle between the current state and the goal.
Unverifiable source (2026) 1 of 2DelegateChoose a list of delegatesto vote as the majority of them.Unless you vote directly.
More
h/ai-safety
votes