Require human-in-the-loop oversight for agentic AI systems acting in high-stakes domains

ai-safety ai-governance policy ethics ai-regulation regulations

Share this with your representatives, journalists and friends

Arguments for

Human final say

Supporters say AI can assist in consequential settings, but decisions affecting life, rights, public services, or legal outcomes should not be made solely by systems. A human should keep the final word, remain accountable, and be able to explain, contest, or halt the outcome when necessary.

However capable of our systems become, a person must keep the final word to understand the decision, to contest it, and where it matters most, to switch the systems off. This is non-negotiable, above all in high-risk systems where a single decision can cost a life, a livelihood, …

Rafal Kownacki · Jul 7, 2026 · full quote
The bill would require the Defense Department to ensure any intentionally lethal use of an autonomous or AI-enabled weapon system is subject to “human oversight, approval, or a human-in-the-loop.”

Human Authority over Autonomous Weapons Act · Jul 17, 2026 · full quote
AI has enormous potential in healthcare, but it cannot replace physician judgment. Patients deserve care decisions that are informed by the latest medical evidence and guided by a physician who understands their individual needs. Whether AI is helping a physician make a clinical …

John Whyte · Jun 10, 2026 · full quote
SAs must ensure that decisions that impact the public are not made without oversight by appropriate staff, who make the final decisions. Furthermore, SA users of AI systems must provide meaningful oversight over all decisions made by an AI system and be able to explain the decisi…

New York State Office of Information Technology Services · Jun 24, 2026 · full quote
High-impact automated decisions, such as customs targeting, identity verification, or logistics prioritization, must remain under clear human accountability and be traceable and contestable.

Universal Postal Union · Jul 7, 2026 · full quote
Humans must retain final control over decisions supported or affected by AI. This is particularly important when AI systems can affect access to rights, public services, employment opportunities, education, healthcare, credit, information, or democratic participation.

Government of Brazil · Jul 7, 2026 · full quote

Kill switches and overrides

A major argument is that increasingly autonomous systems can malfunction, resist intervention, or cause severe harm before people can react. Supporters therefore call for reliable kill switches, override controls, and shutdown authority that do not depend on the AI's cooperation.

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 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
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
Implement accessible kill switches, override controls or independent shutdown mechanisms for every autonomous or safety-relevant edge AI system. These controls should not rely on the AI’s cooperation and should be tested regularly. Define where human approval is required before A…

Canadian Centre for Cyber Security · Jul 15, 2026 · full quote
Because we still understand so little about the inner workings of these systems, ensuring that these more powerful agents stay under human control, that we can correct or stop them if they start behaving badly, is an urgent and persistent problem.

International AI Safety Report · Jul 7, 2026 · full quote

Documented accountability

Many quotes frame human oversight as an operational system, not just a principle. They call for audit trails, monitoring, access controls, documentation, incident reporting, and escalation paths so organizations can see who authorized an action, track what the agent did, and show that oversight was meaningful when it occurred.

Make the technical controls real. That means AI-ready data, clear access controls, audit trails, model and agent monitoring, defined escalation paths and the ability to intervene when something behaves outside expectations.

Joe Depa · Aug 7, 2026 · full quote
An "audit trail should preserve both identities: the human who initiated or authorized the action and the agent that executed it."

Vinay Patel · Jul 29, 2026 · full quote
Material oversight decisions should be documented in a manner proportionate to risk, including the responsible person, information considered, deviations from AI output, overrides, escalations, and resulting action where appropriate.

I Am J Corporation · Jul 14, 2026 · full quote
Responsible deployment requires identity and access controls, data lineage evaluation, bias security testing, red teaming tools, continuous monitoring, transparent incident reporting, audit logs, human oversight, and appealing mechanisms with the ability to shut a system down whe…

Sergio Gago · Jun 23, 2026 · full quote
The FCA’s own use of AI remains subject to the same expectations it puts on firms, including explainability, auditability and appropriate human oversight.

Financial Conduct Authority · Jul 6, 2026 · full quote
For agentic AI, governance cannot depend only on written policies, user disclosures, or generic approval buttons. It needs evidence that oversight was meaningful when the action occurred.

Rishabh Banga · Jun 10, 2026 · full quote

Capable human reviewers

Several quotes argue that oversight works only if the human reviewer is genuinely equipped to judge the system: trained, expert, independent, and given time and authority to challenge outputs. Without that, people may defer to systems they cannot interrogate, and review becomes superficial.

We conclude that effective use of agentic AI requires both domain expertise to frame questions and evaluate outputs, and coding literacy to guide the AI's approach.

Edmund Seto · Aug 7, 2026 · full quote
Governance must ensure that commanders have sufficient training, time and authority to critically evaluate AI recommendations, and that human decision makers retain careful, independent judgment.

Branka Marijan · Aug 4, 2026 · full quote
A health worker trained to trust an AI system they can’t interrogate is not empowered, leading to mistakes outside their control.

Hans Henri P. Kluge · Jul 15, 2026 · full quote
Our approach is grounded in the high-stakes nature of qualifications and the significant impact that results have on students’ progression and life chances. [...] This includes ensuring appropriate expert human involvement and oversight, quality assurance and accountability for o…

Ofqual · Jul 16, 2026 · full quote
Human oversight of AI systems cannot be just a rubber stamp. It requires that an identified person be granted authority, competence, time, independence, and power to alter and even stop a system.

Volker Türk · Jul 7, 2026 · full quote
Humans remain fully accountable for accuracy, appropriateness, ethical use, and final decisions. AI output must be reviewed, edited, and validated before use.

Western Michigan University Homer Stryker M.D. School of Medicine · Jul 11, 2026 · full quote

Arguments against

Velocity and scale limits

Critics argue that routing agent actions to humans does not scale in production systems making many fast decisions. In high-velocity settings, they say human-in-the-loop can slow operations, fail to deliver the intended results, and become a poor fit for autonomous systems acting continuously at scale.

We’re not huge fans of human-in-the-loop. It’s something that you should use judiciously, where you absolutely need it. But it’s not something that you can do at high velocity. You will not get the results that you want to get.

Eric Brandwine · Jun 20, 2026 · full quote
The industry’s current answer is unsatisfying: Human-in-the-Loop (HITL). In development environments and low-frequency pipelines, routing uncertain decisions to a human can be defensible. In production systems operating at scale—dozens of agents, hundreds of decisions per hour—it…

Artur Huk · May 11, 2026 · full quote
Industry discourse often presents “Human-in-the-Loop” as the ultimate safeguard for AI deployment, yet this is a dangerous fallacy in high-velocity agentic systems.

Steve Tidwell · May 1, 2026 · full quote
Our frameworks were not built to contemplate autonomous agents, and relying on a human in the loop for all agent actions is unlikely to be realistic. More sophisticated governance and accountability frameworks may be needed.

Sarah Breeden · Jul 2, 2026 · full quote

Avoid blanket mandates

Another objection is to blanket mandates. These quotes argue that user intervention should scale with the potential impact of an agent's actions and that autonomous agents may need broader governance and accountability frameworks than requiring a human in the loop for all actions.

Policymakers should avoid blanket oversight mandates; instead, recognize that user intervention should scale with the potential impact of the agent’s actions.

Google · May 2026 · full quote
Our frameworks were not built to contemplate autonomous agents, and relying on a human in the loop for all agent actions is unlikely to be realistic. More sophisticated governance and accountability frameworks may be needed.

Sarah Breeden · Jul 2, 2026 · full quote
We’re not huge fans of human-in-the-loop. It’s something that you should use judiciously, where you absolutely need it. But it’s not something that you can do at high velocity. You will not get the results that you want to get.

Eric Brandwine · Jun 20, 2026 · full quote

Middle ground

Graduated oversight models

Some supportive voices favor a graduated approach rather than literal review of every step. They distinguish human-in-the-loop, human-on-the-loop, and human-in-command, and say oversight should intensify with autonomy and consequence, with mandatory approval focused on designated high-impact actions.

Human oversight should increase as agent autonomy and consequence increase.

PwC · Jul 17, 2026 · full quote
To protect employees and exact real productivity, businesses must build in human oversight at the right level – not at every level.

Sonali Fenner · Jul 20, 2026 · full quote
‘Human-in-the-loop’: human approval of all decisions made by AI, which might be appropriate in situations where accuracy is more important than speed of decision-making (e.g. approval of a health insurance claim). [...] ‘Human-in-command’: high-level human oversight, including de…

Financial Stability Board · Jun 10, 2026 · full quote
Conventional oversight models are increasingly insufficient. The shift is moving from “human-in-the-loop” to “human-on-the-loop”, reflecting a more supervisory rather than directly interventionist role for people.

Pieter Danhieux · Jul 13, 2026 · full quote
Organizations should also define where human intervention is required and who is responsible for the decisions models make. Oversight should focus on the activities that carry the greatest operational, financial or regulatory risk.

Vishal Salvi · Aug 7, 2026 · full quote
Our national AI policy ensures human in the loop where AI must support and not replace human decision-making. Human on the loop, where human oversight remains an inbuilt feature. And human in command, where ultimate legal responsibility remains that of humans.

Shahza Fatima Khawaja · Jul 7, 2026 · full quote

Insights

Several supportive quotes say oversight must be more than a formal sign-off: the human reviewer needs authority, competence, time, independence, and the power to change or stop outcomes (12033, 12542, 16186, 14073).
Supporters often pair human oversight with technical controls such as audit trails, access limits, monitoring, documentation, and independent shutdown mechanisms (17363, 17056, 17661, 12435, 17612).
The domains named as high-stakes span defense, justice, healthcare, education, finance, public services, and critical infrastructure, but the recurring concern is the same: protecting life, rights, and accountability (14284, 12036, 17286, 16181, 13658, 11595).
Several supportive quotes already reject one-size-fits-all review, proposing oversight that scales with autonomy and consequence and sometimes shifts from direct approval to supervisory or in-command roles (15683, 17526, 10047, 17362, 12105).
The sharpest disagreement is over feasibility and specificity: some quotes want an explicit human-in-the-loop requirement at critical points, while the opposing quotes warn against making that a blanket rule for every agent action (16187, 11595, 13819, 14125, 15861).

Many quotes emphasize keeping humans in control when agentic AI operates in consequential settings, but there is substantial disagreement over whether that control must take the form of direct human-in-the-loop approval. The central fault line is between advocates of explicit final human authority, documentation, and shutdown powers, and critics who say blanket human-in-the-loop requirements are impractical for fast systems operating at scale and should give way to more proportional supervisory governance.

Cast your vote:

Results (480 votes):

Total (480 votes)

For 467 (97%) Abstain 0 (0%) Against 13 (3%)

For (464)

Edmund Seto
Environmental health researcher and author of a 2026 paper on human-in-the-loop agentic AI.

I am this person! Add a sourced quote Report comment

votes For and says:

We conclude that effective use of agentic AI requires both domain expertise to frame questions and evaluate outputs, and coding literacy to guide the AI's approach.
AI Verified source (Aug 7, 2026)

added 9h ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Joe Depa

I am this person! Add a sourced quote Report comment

votes For and says:

Make the technical controls real. That means AI-ready data, clear access controls, audit trails, model and agent monitoring, defined escalation paths and the ability to intervene when something behaves outside expectations.
AI Verified source (Aug 7, 2026)

added 3d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Vishal Salvi

I am this person! Add a sourced quote Report comment

votes For and says:

Organizations should also define where human intervention is required and who is responsible for the decisions models make. Oversight should focus on the activities that carry the greatest operational, financial or regulatory risk.
AI Verified source (Aug 7, 2026)

added 3d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Rodrigo Coelho
Chief Executive Officer of Edge & Node, writing on AI accountability and security.

I am this person! Add a sourced quote Report comment

votes For and says:

Embedding infrastructural policy enforcement, audit trails, and human-reviewable records does bring an additional layer prior to deployment. And in a domain where nefarious actors operate free of this operational layer, it's fair to ask whether enter...
more AI Verified source (Aug 6, 2026)

added 18h ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Daniel Card

I am this person! Add a sourced quote Report comment

votes For and says:

Security comes from the details; it comes from care. More care needs to go into systems – even more so when they are autonomous. We can't just set them off and hope they are ok. They need monitoring.
AI Verified source (Aug 5, 2026)

added 3d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Branka Marijan
Senior fellow at the Centre for International Governance Innovation and senior researcher at Project Ploughshares.

I am this person! Add a sourced quote Report comment

votes For and says:

Governance must ensure that commanders have sufficient training, time and authority to critically evaluate AI recommendations, and that human decision makers retain careful, independent judgment.
AI Verified source (Aug 4, 2026)

added 2d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Vinay Patel
Chief Security Officer at Zendesk, specializing in enterprise and agentic-AI security.

I am this person! Add a sourced quote Report comment

votes For and says:

An "audit trail should preserve both identities: the human who initiated or authorized the action and the agent that executed it."
AI Verified source (Jul 29, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Edward Chen
Chief AI Officer at NCS

I am this person! Add a sourced quote Report comment

votes For and says:

Organisations should not govern agents like software modules. They should onboard them like digital workers – with clear roles, access limits, supervision and accountability.
AI Verified source (Jul 28, 2026)

added 18h ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Ceren Kaya Akgün
Author and practitioner writing on human-in-the-loop controls for agentic AI workflows.

I am this person! Add a sourced quote Report comment

votes For and says:

Human in the loop for AI agents is an execution-control pattern in which an agent stops before a designated action, hands a decision to a person, and does not continue until that person accepts, edits, or refuses it.
AI Verified source (Jul 25, 2026)

added 15d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Aspen University
Accredited U.S. university and publicly traded higher-education institution.

I am this person! Add a sourced quote Report comment

votes For and says:

AI should not serve as the sole basis for decisions that materially affect a student’s academic standing, progression, financial status, enrollment status, access to university services, disability accommodations, student conduct outcomes, or other s...
more AI Verified source (Jul 24, 2026)

added 15d 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 17d 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) 1 of 2

added 17d 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 17d 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) 1 of 2

added 17d 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 17d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Marlon Oliver
SVP of EMEA & APAC Operations at Flexera

I am this person! Add a sourced quote Report comment

votes For and says:

The organizations that navigate agentic AI well will be the ones that treat governance as part of the deployment decision rather than an afterthought to it.
AI Verified source (Jul 21, 2026)

added 18h ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Shelly Bruce
CIGI distinguished fellow and former chief of Canada's Communications Security Establishment.

I am this person! Add a sourced quote Report comment

votes For and says:

These tools must themselves be held to account: evaluated before they are trusted, governed so they do not widen the attack surface, and kept under meaningful human oversight for consequential decisions.
AI Verified source (Jul 21, 2026)

added 1d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Shannon Williams
Author at Obot writing on enterprise AI governance and Model Context Protocol (MCP) security.

I am this person! Add a sourced quote Report comment

votes For and says:

None of this replaces human review for the decisions that warrant it.
AI Verified source (Jul 21, 2026)

added 20d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Sonali Fenner
Managing Director at Slalom, writing on AI-enabled work and organizational transformation.

I am this person! Add a sourced quote Report comment

votes For and says:

To protect employees and exact real productivity, businesses must build in human oversight at the right level – not at every level.
AI Verified source (Jul 20, 2026)

added 20d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Maidstone Borough Council
Local government authority in Kent, England.

I am this person! Add a sourced quote Report comment

votes For and says:

A serious breach, such as wilfully or negligently uploading confidential personal data to a prohibited service or bypassing human oversight, may constitute gross misconduct and will be investigated according to the Council’s disciplinary procedures.
AI Verified source (Jul 17, 2026)

added 16d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Tom Barrett
U.S. Representative for Michigan's 7th congressional district.

I am this person! Add a sourced quote Report comment

votes For and says:

Artificial intelligence is playing an increasingly critical role in strengthening our national defense, but it should never be the one making life-or-death decisions. We must establish safeguards now to ensure these technologies are used ethically an...
more AI Verified source (Jul 17, 2026)

added 20d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
PwC
Global professional services network publishing research on artificial intelligence and the labor market.

I am this person! Add a sourced quote Report comment

votes For and says:

Human oversight should increase as agent autonomy and consequence increase.
AI Verified source (Jul 17, 2026)

added 20d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Don Beyer
U.S. Representative from Virginia's 8th district; co-sponsor of the GUARDRAILS Act on AI

I am this person! Add a sourced quote Report comment

votes For and says:

No machine should ever be given the power to decide to kill a human being on its own.
AI Verified source (Jul 17, 2026)

added 20d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Lucy Poole
Deputy CEO for Strategy, Planning and Performance at Australia’s Digital Transformation Agency.

I am this person! Add a sourced quote Report comment

votes For and says:

Where a system uses personal information or affects the public, safeguards need to be designed in from the start: human judgement, contestability and proper records.
AI Verified source (Jul 17, 2026)

added 20d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Xi Jinping
President of the People's Republic of China.

I am this person! Add a sourced quote Report comment

votes For and says:

We should put in place laws and regulations, technological monitoring, early warning and emergency response systems in order to strengthen the line of security, prevent abuses and malicious use, and ensure that AI is always under human control.
AI Verified source (Jul 17, 2026)

added 23d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Sofia Guerra
Government Relations Associate at Win Without War, an advocacy organization focused on U.S. foreign policy and military accountability.

I am this person! Add a sourced quote Report comment

votes For and says:

Decisions impacting human life and safety should never be handed off to algorithms.
AI Verified source (Jul 17, 2026)

added 24d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Naveed Shah
Political Director of Common Defense and U.S. Army veteran.

I am this person! Add a sourced quote Report comment

votes For and says:

As an Army veteran who served in Iraq, I know the moral responsibility that comes with taking human life in combat, decisions that must never be handed over to machines.
AI Verified source (Jul 17, 2026)

added 24d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
John Tierney
Executive Director of Council for a Livable World and former U.S. Representative for Massachusetts.

I am this person! Add a sourced quote Report comment

votes For and says:

Every lawmaker should understand the importance of keeping Americans safe by ensuring a human is always involved in decisions affecting human life.
AI Verified source (Jul 17, 2026)

added 24d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Sara Jacobs
U.S. Representative from California's 51st congressional district

I am this person! Add a sourced quote Report comment

votes For and says:

These aren’t risks worth taking – not when they can result in unnecessary lives lost or war. That’s why I’m proud to join Congressman Beyer and Congressman Barrett in introducing legislation to ensure that humans are meaningfully involved before auto...
more AI Verified source (Jul 17, 2026)

added 24d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Human Authority over Autonomous Weapons Act
Bipartisan U.S. federal bill introduced in 2026 to require human control over lethal autonomous weapons.

I am this person! Add a sourced quote Report comment

votes For and says:

The bill would require the Defense Department to ensure any intentionally lethal use of an autonomous or AI-enabled weapon system is subject to “human oversight, approval, or a human-in-the-loop.”
AI Verified source (Jul 17, 2026)

added 24d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate