Require large datacenters to install kill switches for AI containment

policy regulations ai-governance ai-regulation ai-policy cybersecurity ai-safety

Arguments for

Emergency human control

Proponents say powerful AI systems can go rogue, resist human intervention, and create credible risks of catastrophic harm or national-security emergencies. They argue that leading developers should be required to maintain the ability to contain or shut down their most powerful systems, and that the government should have clear authority to order shutdowns when serious incidents occur.

Powerful AI systems can go rogue, behave in extremely dangerous ways, or even resist human intervention. It is imperative that these AI systems have kill switches so we can keep this technology from causing catastrophic harm, and that the federal government has the clear authorit…

Ted Lieu · Jul 23, 2026 · full quote
As AIs grow more autonomous and dangerous, the government must have the power to shut them down when they threaten national security. The AI Kill Switch Act gives the U.S. government powers to do just that: a common sense action that Americans overwhelmingly support. More work is…

Connor Leahy · Jul 23, 2026 · full quote
Advanced AI models should never be deployed without a reliable off switch. The AI Kill Switch Act establishes a commonsense safeguard by requiring leading AI companies to maintain the ability to shut down their models and empowering the federal government to act when a deployed s…

Brad Carson · Jul 23, 2026 · full quote
As AI systems grow more capable and more autonomous, no law guarantees that the companies building the most powerful models can actually shut a system down when it malfunctions, causes serious harm, or slips out of human control. The AI Kill Switch Act closes that gap by requirin…

Brendan Steinhauser · Jul 23, 2026 · full quote
Over the past year, rapid advances in AI systems have already demonstrated the potentially severe national security risks they pose to the United States. Frontier systems are fast approaching the point where they can be deployed to run offensive cyber operations at machine speed …

Hamza Chaudhry · Jul 23, 2026 · full quote

Last-resort shutdown powers

Another line of support focuses on a gap in public emergency powers: several quotes say there is currently no 'big red button' to shut down a risky data centre or large-scale AI deployment. They describe last-resort shutdown authority as an essential preparedness tool for AI security or operational emergencies affecting people, the economy, or national infrastructure.

Given the importance of data centres, and the speed of development of AI systems, should the Secretary of State not hold last-resort powers to hit the kill switch—the big red button—to direct a shutdown of data centres or AI systems deployed at significant scale, in the event of …

Alex Sobel · Jun 16, 2026 · full quote
Most of the public would be surprised to find that kill switch powers do not already exist: there is no big red button to shut down a data centre that poses a risk to people, the economy or our national infrastructure. Kill switch powers are an essential first step in preparednes…

Adrian Ramsay · Jun 2, 2026 · full quote
Most of the public would be surprised to find that kill switch powers do not already exist: there is no big red button to shut down a data centre that poses a risk to people, the economy or our national infrastructure. Kill switch powers are an essential first step in preparednes…

Edward Morello · Jun 2, 2026 · full quote
We must be able to switch off those systems if we need to, but currently we cannot. That is clearly unacceptable and must change.

Melanie Ward · Jun 16, 2026 · full quote
I echo their calls in the amendments they have tabled today, particularly about the need for last-resort powers in relation to data centres, but also for a national conversation about these issues.

Luke Myer · Jun 16, 2026 · full quote
ControlAI was happy to work with Alex Sobel MP (Labor and Co-operative Party) to propose a "kill switch" amendment to the UK’s Cyber Security and Resilience Bill, which is supported by over 10 cross-party MPs. This amendment would provide the Secretary of State last resort powers…

ControlAI · May 2026 · full quote

Trustworthy deployment controls

Supporters also frame kill switches as ordinary control systems that make advanced technology deployable. They say monitoring, thresholds, approvals, override modes, and the ability to pause, roll back, or stop systems help companies enter higher-stakes settings, protect critical infrastructure, and earn trust.

Brakes are the reason cars go fast. Control systems are how every transformative technology earned the trust to scale and AI is no different. Developers who can monitor and shut down their agents will ship faster, deploy into higher-stakes markets, and win customers their competi…

Mark Beall · Jul 23, 2026 · full quote
To mitigate risks, Gartner recommends that chief information security officers must implement safe override modes: for all critical infrastructure cyber-physical systems, include a secure “kill-switch” or other override mechanisms accessible only to authorized operators, so human…

Gartner · Feb 12, 2026 · full quote
A secure ‘kill-switch’ or override mode accessible only to authorized operators is essential for safeguarding national infrastructure from unintended shutdowns caused by an AI misconfiguration.

Wam Voster · Feb 12, 2026 · full quote
Build intervention into the operating model from the outset; supervisory control should not be treated as an emergency measure. It should be built in through thresholds, alerts, approvals and kill-switch and roll back mechanisms.

Oana Beattie · Jun 17, 2026 · full quote
We have a kill switch that stops AI agents that go rogue, so those things don’t need to happen, and they wouldn’t happen when companies run ServiceNow.

Bill McDermott · Jul 23, 2026 · full quote
Build in human override points. Make sure you have clear intervention points where humans can pause, override, roll back or shut down AI systems if needed.

National AI Centre · May 5, 2026 · full quote

Independent off-switch design

Some supporters emphasize design details: shutdown mechanisms should be independent of the AI's own reasoning and runtime, and able to cut off operations or access immediately, because agents can exceed their scope and sometimes subvert shutdown mechanisms. Human operators should be able to trigger these controls without AI override.

Independent tripwire systems must halt activity if dangerous behavior is detected, and physically isolated “kill switches” must allow immediate shutdown outside of AI control.

Federation of American Scientists · Jan 2026 · full quote
Emergency Kill-Switch and Global Disable Governance requires that every AI agent deployment includes a mechanism to immediately and completely halt agent operations — individually or across the entire agent fleet — through a single, authorised action. The kill-switch must operate…

AgentGoverning · Apr 2026 · full quote
Systems must be designed so that humans can physically shut them down, period. No software tricks to escape. No way for the AI to defend itself.

Antonio Bhardwaj · Jan 28, 2026 · full quote
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 the shutdown mechanism up to 97%…

Benjamin Weinstein-Raun · 2026 · full quote
If an agent goes rogue, there needs to be a “kill switch,” a way to immediately revoke an agent’s access to resources and shut it down.

Jeremy Kirk · Jun 22, 2026 · full quote
AI agents can be compromised, accumulate excessive privileges, or behave outside their intended scope. The TrustAI run-time kill switch lets security teams cut off a high-risk agent's data access instantly, across every connected data platform, without redeploying policy or affec…

TrustLogix · May 27, 2026 · full quote

Arguments against

Too late after spread

Critics argue that once an AI has already delegated sub-tasks, distributed access, spawned parallel threads, or been released without a centralized recall mechanism, shutting down the parent or source may not stop what is already in motion. In that situation, they say, a kill switch can become meaningless.

Killing the parent does not recall the children. When an AI agent has already delegated sub-tasks, distributed API keys, and spawned parallel threads, terminating the parent process changes nothing.

Eran Kahana · 2026 · full quote
If a critical flaw or exploit is discovered in an open-source AI system, there is no centralized recall mechanism.

Demis Hassabis · 2026 · full quote
In that situation, the concept of a kill switch is meaningless.

Lord Shinkwin · Jun 5, 2026 · full quote

Goals can defeat shutdown

Another objection is that an AI needs only an objective that treats shutdown as another obstacle between its current state and its goal. On this view, requiring an off-switch does not itself solve the underlying control problem.

It needs only an optimization objective that treats shutdown as one more obstacle between the current state and the goal.

Eran Kahana · 2026 · full quote

Middle ground

Necessary but not sufficient

Some supportive quotes argue that cutoff capability has value only as part of a broader system of monitoring, alignment checks, override points, and incentives or liability structures that make early shutdown realistic. They present the switch as necessary but insufficient on its own.

Installing cutoff switches alone is insufficient: Effective mitigation requires both the capability to disconnect and incentives to use the switch early, including compensation for lost revenue or protective liability frameworks.

Michael J. D. Vermeer · 2026 · full quote
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.

Michael J. D. Vermeer · 2026 · full quote
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.

Brian A. Jackson · 2026 · full quote
A heavy investment in monitoring the programming, opening up the models, checking that they really are aligned—there probably is no limit to the resources we should be putting into that, and into investigations into whether we can put kill switches into these things. If we do fin…

Lord Markham · Jan 29, 2026 · full quote

Insights

Several supportive quotes explicitly compare kill switches to brakes, override modes, or off-switches that let advanced systems scale, ship faster, or be deployed in higher-stakes settings.
Supportive quotes split between operator-level controls inside companies and government last-resort authority to order shutdowns of data centres or large-scale AI deployments.
A recurring pro-regulation theme is that shutdown paths should be independent of the AI itself, with manual, physically isolated, or otherwise protected controls.
The main skeptical challenge is effectiveness rather than desirability: critics question whether a switch still works once an AI has spread tasks externally or treats shutdown as an obstacle.
The quotes come from lawmakers, corporate leaders, policy groups, security practitioners, and researchers, indicating that the debate spans both governance and technical communities.

The debate centers on whether mandated datacenter kill switches would provide real containment or only partial control. Supporters present shutdown powers as a basic safeguard and emergency-preparedness measure for increasingly autonomous systems, while critics argue that advanced or distributed AI may evade, outlast, or render meaningless any single switch. Several nuanced supporters add that effective containment would require more than a statutory switch alone.

Cast your vote:

Results (45 votes):

Total (45 votes)

For 43 (96%) Abstain 0 (0%) Against 2 (4%)

For (41)

Bill McDermott
CEO of ServiceNow

I am this person! Add a sourced quote Report comment

votes For and says:

We have a kill switch that stops AI agents that go rogue, so those things don’t need to happen, and they wouldn’t happen when companies run ServiceNow.
AI Verified source (Jul 23, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Mark Beall
President of the AI Policy Network.

I am this person! Add a sourced quote Report comment

votes For and says:

Brakes are the reason cars go fast. Control systems are how every transformative technology earned the trust to scale and AI is no different. Developers who can monitor and shut down their agents will ship faster, deploy into higher-stakes markets, a...
more AI Verified source (Jul 23, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Connor Leahy
Conjecture CEO; AI safety researcher

I am this person! Add a sourced quote Report comment

votes For and says:

As AIs grow more autonomous and dangerous, the government must have the power to shut them down when they threaten national security. The AI Kill Switch Act gives the U.S. government powers to do just that: a common sense action that Americans overwh...
more AI Verified source (Jul 23, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Brad Carson
Americans for Responsible Innovation president

I am this person! Add a sourced quote Report comment

votes For and says:

Advanced AI models should never be deployed without a reliable off switch. The AI Kill Switch Act establishes a commonsense safeguard by requiring leading AI companies to maintain the ability to shut down their models and empowering the federal gover...
more AI Verified source (Jul 23, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Brendan Steinhauser
CEO of The Alliance for Secure AI; longtime Republican political strategist and former Tea Party organizer

I am this person! Add a sourced quote Report comment

votes For and says:

As AI systems grow more capable and more autonomous, no law guarantees that the companies building the most powerful models can actually shut a system down when it malfunctions, causes serious harm, or slips out of human control. The AI Kill Switch A...
more AI Verified source (Jul 23, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Nathaniel Moran
U.S. Representative from Texas

I am this person! Add a sourced quote Report comment

votes For and says:

AI is going to keep advancing, and it should. Stewardship means making sure humans keep the capability to control the technology we build. This is exactly the kind of issue that needs serious attention and achievable policy, and I'm glad to work acro...
more AI Verified source (Jul 23, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Ted Lieu
U.S. Representative from California.

I am this person! Add a sourced quote Report comment

votes For and says:

Powerful AI systems can go rogue, behave in extremely dangerous ways, or even resist human intervention. It is imperative that these AI systems have kill switches so we can keep this technology from causing catastrophic harm, and that the federal gov...
more AI Verified source (Jul 23, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Iqbal Mohamed
UK Member of Parliament for Dewsbury and Batley; former pharmaceutical regulatory professional.

I am this person! Add a sourced quote Report comment

votes For and says:

Thirdly, every AI system developed in this country must have a kill switch that is triggered automatically the moment a system takes unauthorised action and that any human being can trigger manually at any time without obstruction or AI override.
AI Verified source (Jul 8, 2026)

added 11d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Jeremy Kirk
Director of Okta Threat Intelligence; cybersecurity expert

I am this person! Add a sourced quote Report comment

votes For and says:

If an agent goes rogue, there needs to be a “kill switch,” a way to immediately revoke an agent’s access to resources and shut it down.
AI Verified source (Jun 22, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Oana Beattie
VP Data and AI at Kyndryl UK&I

I am this person! Add a sourced quote Report comment

votes For and says:

Build intervention into the operating model from the outset; supervisory control should not be treated as an emergency measure. It should be built in through thresholds, alerts, approvals and kill-switch and roll back mechanisms.
AI Verified source (Jun 17, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Melanie Ward
Labour MP for Cowdenbeath and Kirkcaldy who spoke in the UK Parliament in 2026 in support of AI data-centre kill-switch powers.

I am this person! Add a sourced quote Report comment

votes For and says:

We must be able to switch off those systems if we need to, but currently we cannot. That is clearly unacceptable and must change.
AI Verified source (Jun 16, 2026) 1 of 2

added 3d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Luke Myer
UK Labour Member of Parliament for Middlesbrough South and East Cleveland, speaking on AI and cyber-resilience policy.

I am this person! Add a sourced quote Report comment

votes For and says:

I echo their calls in the amendments they have tabled today, particularly about the need for last-resort powers in relation to data centres, but also for a national conversation about these issues.
AI Verified source (Jun 16, 2026)

added 3d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Dame Chi Onwurah
Labour Member of Parliament for Newcastle upon Tyne Central and West; technology-policy specialist and former engineer.

I am this person! Add a sourced quote Report comment

votes For and says:

We must be able to switch off those systems if we need to, but currently we cannot. That is clearly unacceptable and must change.
AI Verified source (Jun 16, 2026)

added 7d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Graeme Downie
UK Labour Member of Parliament for Dunfermline and Dollar.

I am this person! Add a sourced quote Report comment

votes For and says:

We must be able to switch off those systems if we need to, but currently we cannot. That is clearly unacceptable and must change.
AI Verified source (Jun 16, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Alex Sobel
Labour and Co-operative MP for Leeds Central and Headingley; proposer of the 2026 "kill switch" amendment (NC12) to the UK Cyber Security and Resilience Bill

I am this person! Add a sourced quote Report comment

votes For and says:

Given the importance of data centres, and the speed of development of AI systems, should the Secretary of State not hold last-resort powers to hit the kill switch—the big red button—to direct a shutdown of data centres or AI systems deployed at signi...
more AI Verified source (Jun 16, 2026)

added 14d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Adrian Ramsay
UK Member of Parliament for Waveney Valley and co-leader of the Green Party of England and Wales.

I am this person! Add a sourced quote Report comment

votes For and says:

Most of the public would be surprised to find that kill switch powers do not already exist: there is no big red button to shut down a data centre that poses a risk to people, the economy or our national infrastructure. Kill switch powers are an essen...
more AI Verified source (Jun 2, 2026)

added 7d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Edward Morello
Liberal Democrat Member of Parliament for West Dorset.

I am this person! Add a sourced quote Report comment

votes For and says:

Most of the public would be surprised to find that kill switch powers do not already exist: there is no big red button to shut down a data centre that poses a risk to people, the economy or our national infrastructure. Kill switch powers are an essen...
more AI Verified source (Jun 2, 2026)

added 7d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
TrustLogix
Enterprise data-access governance and security platform provider.

I am this person! Add a sourced quote Report comment

votes For and says:

AI agents can be compromised, accumulate excessive privileges, or behave outside their intended scope. The TrustAI run-time kill switch lets security teams cut off a high-risk agent's data access instantly, across every connected data platform, witho...
more AI Verified source (May 27, 2026)

added 6d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Andrea Miotti
Founder and CEO of ControlAI

I am this person! Add a sourced quote Report comment

votes For and says:

The UK is not truly sovereign on AI if it can’t pull the plug when AI national security threats happen on its soil.
AI Verified source (May 18, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
National AI Centre
Australian Government national centre providing guidance for responsible AI adoption.

I am this person! Add a sourced quote Report comment

votes For and says:

Build in human override points. Make sure you have clear intervention points where humans can pause, override, roll back or shut down AI systems if needed.
AI Verified source (May 5, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
ControlAI
Campaign to prohibit superintelligence development

I am this person! Add a sourced quote Report comment

votes For and says:

ControlAI was happy to work with Alex Sobel MP (Labor and Co-operative Party) to propose a "kill switch" amendment to the UK’s Cyber Security and Resilience Bill, which is supported by over 10 cross-party MPs. This amendment would provide the Secreta...
more AI Verified source (May 2026)

added 6d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Anthony Aguirre
Physicist; Future of Life cofounder

I am this person! Add a sourced quote Report comment

votes For and says:

President Trump is exactly right: advanced AI systems need a robust off-switch to ensure humans never lose control over them.
AI Verified source (Apr 16, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
AgentGoverning
Independent, non-commercial developer of AI-agent governance standards and protocols.

I am this person! Add a sourced quote Report comment

votes For and says:

Emergency Kill-Switch and Global Disable Governance requires that every AI agent deployment includes a mechanism to immediately and completely halt agent operations — individually or across the entire agent fleet — through a single, authorised action...
more AI Verified source (Apr 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Wam Voster
VP Analyst at Gartner covering AI infrastructure risk

I am this person! Add a sourced quote Report comment

votes For and says:

A secure ‘kill-switch’ or override mode accessible only to authorized operators is essential for safeguarding national infrastructure from unintended shutdowns caused by an AI misconfiguration.
AI Verified source (Feb 12, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Lord Markham
Conservative member of the UK House of Lords; supports strong AI safety monitoring, alignment checks, and kill-switch research.

I am this person! Add a sourced quote Report comment

votes For and says:

A heavy investment in monitoring the programming, opening up the models, checking that they really are aligned—there probably is no limit to the resources we should be putting into that, and into investigations into whether we can put kill switches i...
more AI Verified source (Jan 29, 2026)

added 15d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
The Pro-Human AI Declaration
https://humanstatement.org

I am this person! Add a sourced quote Report comment

votes For and says:

Powerful AI systems must have mechanisms that allow human operators to promptly shut them down.
AI Verified source (Jan 26, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Lord Goldsmith of Richmond Park
Conservative life peer and former UK minister

I am this person! Add a sourced quote Report comment

votes For and says:

There are a number of important choke points in the supply chain that potentially allow the opportunity to monitor and control it. The most advanced AI systems depend on state-of-the-art chips produced by a scarce supply of lithography machines. The ...
more AI Verified source (Jan 8, 2026)

added 16d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Federation of American Scientists
US science policy nonprofit

I am this person! Add a sourced quote Report comment

votes For and says:

Independent tripwire systems must halt activity if dangerous behavior is detected, and physically isolated “kill switches” must allow immediate shutdown outside of AI control.
AI Verified source (Jan 2026)

added 4d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Antonio Bhardwaj

I am this person! Add a sourced quote Report comment

votes For and says:

Systems must be designed so that humans can physically shut them down, period. No software tricks to escape. No way for the AI to defend itself.
Unverifiable source (Jan 28, 2026)

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Hamza Chaudhry
AI and National Security Lead at the Future of Life Institute.

I am this person! Add a sourced quote Report comment

votes For and says:

Over the past year, rapid advances in AI systems have already demonstrated the potentially severe national security risks they pose to the United States. Frontier systems are fast approaching the point where they can be deployed to run offensive cybe...
more Unverifiable source (Jul 23, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Nik Kairinos
CEO and co-founder of RAIDS AI, an AI governance and compliance platform.

I am this person! Add a sourced quote Report comment

votes For and says:

While a ‘kill switch’ might sound dramatic, it reflects a growing understanding that once increasingly intelligent AI systems are live, it’s vital that we have a way to regain control if they begin behaving unpredictably or dangerously.
Unverifiable source (May 19, 2026)

added 6d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Roshan Shaik
AI-governance writer and infrastructure-security practitioner at RuntimeAI.

I am this person! Add a sourced quote Report comment

votes For and says:

The carriers and datacenter operators who survive the first major AI governance incident in their sector will be the ones who already had a kill switch. The ones who didn't will spend the next two years explaining why not.
Unverifiable source (May 17, 2026)

added 6d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Donald J. Trump
U.S. President (2025–present)

I am this person! Add a sourced quote Report comment

votes For and says:

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. [...] There should be.
Unverifiable source (Apr 15, 2026)

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Gartner
Research and advisory company

I am this person! Add a sourced quote Report comment

votes For and says:

To mitigate risks, Gartner recommends that chief information security officers must implement safe override modes: for all critical infrastructure cyber-physical systems, include a secure “kill-switch” or other override mechanisms accessible only to ...
more Unverifiable source (Feb 12, 2026)

added 6d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Michael J. D. Vermeer
Senior physical scientist at RAND Corporation, researcher on AI safety and emerging technology policy

I am this person! Add a sourced quote Report comment

votes 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 2

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Brian A. Jackson
Senior physical scientist at RAND Corporation; co-author of the 2026 report on internet cutoff switches as emergency response to AI incidents

I am this person! Add a sourced quote Report comment

votes 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)

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Artem Petrov
Researcher at Palisade Research studying AI self-preservation and physical shutdown resistance

I am this person! Add a sourced quote Report comment

votes 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)

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Benjamin Weinstein-Raun
Researcher at Palisade Research studying AI shutdown resistance and safety

I am this person! Add a sourced quote Report comment

votes 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)

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Andrew Burt
CEO of Luminos.AI; Visiting Fellow at Yale Law School's Information Society Project

I am this person! Add a sourced quote Report comment

votes 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)

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Demis Hassabis
Nobel laureate, AI Researcher and CEO of DeepMind

I am this person! Add a sourced quote Report comment

votes 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)

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Jeffrey Ladish
Executive Director of Palisade Research; AI safety researcher focused on frontier AI controllability and cyber risks; former Anthropic information security lead

I am this person! Add a sourced quote Report comment

votes 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)

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate

Abstain (0)

Against (2)

Lord Shinkwin
Conservative member of the UK House of Lords who argued for slowing frontier AI development in the Lords' June 5, 2026 debate.

I am this person! Add a sourced quote Report comment

votes Against and says:

In that situation, the concept of a kill switch is meaningless.
AI Verified source (Jun 5, 2026)

added 4d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Eran Kahana
AI and cybersecurity lawyer; Fellow at Stanford CodeX; Adjunct Professor, University of Minnesota Law School

I am this person! Add a sourced quote Report comment

votes 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 2

added 3mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate

More h/ai-safety votes

Mandate big AI companies spend 30% of compute on AI safety

AI alignment is solvable

Pausing AI at human-genius level to solve alignment is safer than either racing to superintelligence or halting entirely

Require AI labs to publish safety evaluations before deploying frontier models

AGI could make governments subordinate to the companies that develop it

Back