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

Key insights

Many quotes define high-stakes use by effects on safety, rights, finances, employment, healthcare, public services, or critical infrastructure.
Supporters usually present human oversight as one layer in a broader control stack that also includes identity, access limits, testing, monitoring, logging, audit trails, escalation paths, and shutdown or override mechanisms.
Several supportive quotes reject constant human micromanagement and instead advocate risk-proportional or graduated oversight tied to consequence, reversibility, or blast radius.
The clearest opposition focuses on feasibility and response time rather than rejecting governance outright, arguing that some environments need controls that act faster than human approval can.
Across government, corporate, academic, and international voices, a recurring theme is that humans must remain accountable and able to change or stop outcomes, not merely observe them.

Arguments for

Irreversible actions need approval

Proponents say actions that are high-impact, irreversible, or capable of causing a large blast radius should stop for explicit human approval before they happen. They point to steps such as moving money, changing production systems, releasing sensitive data, or affecting safety or customer eligibility as decisions that should not proceed autonomously.

Require fresh human approval before an irreversible or high-impact action, then log the identity, request, policy decision, approval, and resulting change in a system the agent cannot alter.

iTechGuides Team · Oct 18, 2026 · full quote
The trick is moving away from constant human micromanagement and adopting risk-proportional oversight instead. Low-risk tasks like summarising internal documents can run fully automated, while medium-risk tasks like updating database records can execute automatically but queue up…

Bernard Montel · Oct 11, 2026 · full quote
An agent is a set of actions: read a ticket, look up a customer, draft a reply, update a field, send an email, issue a refund. Each action carries different risk. The right question is: What action needs what kind of human approval? [...] Low reversibility or a large blast radius…

Dipankar Sarkar · Oct 8, 2026 · full quote
Human approval should remain mandatory for actions that could jeopardize human safety, create material financial loss, change production systems, delete or disclose sensitive data, alter security policies, create legal commitments, affect employment or significantly impact custom…

Mohan Veloo · Sep 7, 2026 · full quote
An agent that can reconcile invoices, route support tickets or prepare a management report may save time with less risk. The same agent should not be allowed to approve a large payment, change a production system, make a hiring decision or send an external commitment without huma…

Tamaghn Kasibhatta · Oct 6, 2026 · full quote
They should not allow high stakes or irreversible actions to take place without human review. Appropriate actions therefore include identifying checkpoints or action boundaries that require human approval.

Ministry of Digital Development and Information (Singapore) · May 6, 2026 · full quote

Human judgment for rights

Supporters argue that when AI affects rights, livelihoods, public services, employment, healthcare, legal outcomes, or lethal force, human judgment must remain decisive because responsibility cannot be delegated to a system. They emphasize identifiable accountable people, contestability, and redress rather than allowing institutions to defer to automated outputs.

As AI becomes more autonomous the real question for business leaders is where human judgment matters most. What should machines be allowed to do independently? What should trigger human intervention? And who remains accountable for the outcome? [...] At the highest level are deci…

Alan Turnley-Jones · Sep 22, 2026 · full quote
If we want to strengthen our justice systems, I am of the view that we have to focus on four priorities: to strengthen institutional capacities of lawyers and jurists, to look at transparency and traceability in the use of AI, to have human oversight that is effective, and to ens…

Samuel Arsenault · Jul 7, 2026 · full quote
We need to appoint an identifiable human being in charge for all issues that are tied to rights, and there needs to be a right to redress in good time, in line with the speed of AI systems, so that humans can receive redress if necessary. Therefore, the functioning of the algorit…

Amal El Fallah Seghrouchni · 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.

Eugênio Garcia · Jul 7, 2026 · full quote
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 autonomous or AI-enabled weapon sy…

Sara Jacobs · Jul 17, 2026 · full quote
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

Unpredictable systems need oversight

Another argument is that current agentic systems are still too hard to understand, too error-prone on complex tasks, and too capable of acting across sensitive systems to be trusted alone in high-stakes settings. Because their behavior cannot always be reliably explained or safely predicted, humans should retain governance, review, and stopping power.

That same kind of mechanistic understanding eludes us in today’s Super Intelligence systems, even as the frontier models powering these systems are now more capable than traditional software systems. We can’t attribute model behaviors and outputs to specific inputs of training da…

Satya Nadella · Oct 12, 2026 · full quote
State-of-the-art AI agents can have error rates exceeding 25 per cent on complex tasks, a situation that is unacceptable in high-stakes decision-making. Organisations seeking accuracy and repeatability must embed domain expertise into AI-agent workflows while keeping humans at th…

Bryan Harris · Oct 2, 2026 · full quote
As organizations unleash the power of agent autonomy, it raises the pivotal and non-negotiable requirement of enforcing governance. [...] With basic AI systems, risk is much more contained compared to that of an autonomous agent that has access to multiple business systems and ca…

Mahmoud Ramin · Oct 9, 2026 · full quote
The complexity and unpredictability of real world scenarios mean that even the most sophisticated AI agent can encounter uncertainty, ambiguous requirements, and high-stakes decisions that are beyond its capabilities. This is where human oversight becomes essential.

Oracle · Jul 13, 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
because AI agents can act directly upon the real world, virtually and increasingly in the physical world, their failures can potentially cause severe harm before a human can even intervene.

Concordia AI · Jul 7, 2026 · full quote

Governance must be built in

Proponents often frame oversight as a design requirement rather than an afterthought: organizations should define what agents may do autonomously, give them limited identities and permissions, monitor actions, keep audit trails, and build approval workflows and override mechanisms from the start. In this view, bounded autonomy can speed low-risk work while keeping accountable leaders in control of material consequences.

The future model is policy-governed autonomy: AI detects, prioritizes and executes approved low-risk actions at machine speed, while accountable leaders retain control of decisions with material operational, legal, regulatory or reputational consequences. The goal is faster actio…

Kapil Tiwari · Oct 16, 2026 · full quote
Governance must be designed into the deployment as well. Organizations should define which actions systems may take autonomously, when human review is required and how decisions can be understood or challenged. The level of oversight should reflect the potential consequences of t…

Meagen Eisenberg · Oct 8, 2026 · full quote
Every AI agent should have its own identity and access to only what it needs. [...] High-impact or irreversible actions should require human approval.

Jaydeep Singh · Oct 17, 2026 · full quote
An agent with production access, payment authority, persistent credentials, or the ability to direct other agents should face stronger evaluation, logging, approval, and incident-reporting expectations. [...] For serious incidents, a report should answer a small set of operationa…

Gleb Tsipursky · Sep 24, 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
Design for human oversight from Day 1: Don't bolt on governance after deployment. Build approval workflows, block states, and audit trails directly into your architecture.

Jaleh Reeves · Jul 1, 2026 · full quote

Arguments against

Scalability and speed limits

Critics argue that human-in-the-loop review should be used judiciously, not as a default requirement, because in production systems with many agents and decisions per hour it may not be realistic or compatible with high velocity. They warn that routing operational decisions to people can become a scalability trap and undermine performance.

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
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

Immediate automated containment

In security settings, some argue that the decisive control is the ability to continuously verify identities and automatically restrict access when trust changes, without waiting on a ticket or human approval. They present immediate containment as more effective than human approval bottlenecks for preventing larger incidents.

A lot of tools can tell you an AI agent is behaving strangely. Very few can act on that at the network layer, without waiting on a human to approve a ticket. That’s the piece we’re closing.

Garrett Gross · Aug 18, 2026 · full quote
Every identity that can connect, access data, or act must be continuously verified and governed. Portnox gives organizations the ability to immediately restrict access when trust changes, without waiting for an AI agent to create a larger security incident.

Denny LeCompte · Aug 18, 2026 · full quote

Middle ground

Risk-based oversight modes

A common nuanced position is that oversight should be graduated rather than uniform: routine tasks can be automated, medium-risk actions may be periodically checked, and high-stakes or irreversible steps trigger direct human review. This approach shifts people away from constant micromanagement and toward supervising boundaries, exceptions, and the highest-consequence decisions.

As agentic AI moves from generating content to making decisions and taking action, boards and C-suite leaders are looking beyond “human in the loop” as their AI oversight model. [...] The newer models don't eliminate the loop. It still belongs in high-stakes or irreversible decis…

John M. Bremen · Oct 16, 2026 · full quote
Human-in-the-loop should be designed as a graduated oversight model rather than a single pattern. The right mode depends on risk, confidence, ambiguity, regulatory sensitivity, operational maturity, and the trustworthiness of the automation.

Florian Lauck-Wunderlich · Jul 24, 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
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 oversight should increase as agent autonomy and consequence increase.

PwC · Jul 17, 2026 · full quote
Deeper human review is concentrated on the AI products that carry the greatest risk.

Steven Mills · Sep 10, 2026 · full quote

Meaningful oversight capacity

Another nuanced view is that oversight must be meaningful, not merely formal. It depends on whether designated people actually have the authority, time, competence, independence, and system support to intervene, with escalation paths, independent checks, and reviewable records built into the system.

Existing responsible AI governance frameworks mandate human oversight. Yet they share a critical assumption: that competent human overseers remain present and capable of exercising meaningful oversight and control over AI decisions. Accountability must be treated as a capacity-se…

Gabriel Sze · Apr 20, 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
Human review should be treated as a decision checkpoint with clear authority, defined escalation paths, and measurable accountability.

NHI Management Group · Jul 9, 2026 · full quote
Its constructive proposal is engineered heterogeneity: the process that proposes an action should not serve as its sole approver and auditor. Independent monitoring and review over time provide additional checks on that process.

Nathan DeBardeleben · Aug 4, 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

The debate is less about whether consequential agentic AI needs governance than about what kind of human control is workable and sufficient. Most quotes support explicit human review, approval, or stopping authority for high-stakes actions and tie that to accountability, rights protection, and safety. The main challenge raised by critics is whether strict human-in-the-loop requirements can scale or respond quickly enough in fast-moving systems, prompting calls for more automated and architectural forms of control.

Cast Your Vote:

Results (562 votes):

Total (562 votes)

For 546 (97%) Abstain 1 (0%) Against 15 (3%)

New Top

For (541)

Maureen Doyle-Spare
Independent researcher in AI governance and founder of Doyle-Spare Research, specializing in agentic AI governance.

I am this person! Add a sourced quote Report comment

votes For and says:

On each in-scope decision, a runtime measurement, the Semantic Deviation Index quantifies how far the agent's resolved meaning sits from the authorized one. A Deterministic Gate turns that measurement into an enforced decision before execution author...
more AI Verified source (Oct 29, 2026)

added 3d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Sandhya Arun
Chief Technology Officer at Wipro, focused on enterprise AI and technology transformation.

I am this person! Add a sourced quote Report comment

votes For and says:

Agentic systems therefore need clear triggers for human review and escalation. During solution design, leaders should define who owns each decision, what agents may and may not do, who can approve or override an action, how exceptions will be escalat...
more AI Verified source (Oct 29, 2026)

added 3d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Husham Ahmed
Counsellor at Pakistan’s Permanent Mission to the United Nations Office and other International Organizations in Geneva.

I am this person! Add a sourced quote Report comment

votes For and says:

Without effective human judgement and control, these systems increase the risks of miscalculation and escalation, lower the threshold for the use of force, and erode compliance with international humanitarian law (IHL). The decision to employ nuclear...
more AI Verified source (Oct 25, 2026)

added 7d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Dipesh Kaura
Country Director, India and SAARC at Securonix, a cybersecurity company.

I am this person! Add a sourced quote Report comment

votes For and says:

As organisations introduce Agentic AI into security operations, human oversight should not be treated as an afterthought. Furthermore, AI should empower human expertise and not replace it. Agentic AI has to be designed with analysts at the center, en...
more AI Verified source (Oct 24, 2026)

added 8d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Shirin Hamid
Chief Information Officer and Director of the Information Technology Department at the International Monetary Fund; digital-governance and emerging-technology leader.

I am this person! Add a sourced quote Report comment

votes For and says:

The intelligent central bank is not the bank with the most AI, It is the bank whose AI can be trusted, and can ensure that any AI that touches the financial system is explainable, governed, secure, ethical, and human-supervised. [...] For every AI-in...
more AI Verified source (Oct 23, 2026)

added 9d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Carolyn Duby
Field CTO and Cybersecurity Lead at Cloudera; cybersecurity, data, and AI expert.

I am this person! Add a sourced quote Report comment

votes For and says:

Organizations should evaluate consequential actions against the agent’s assigned task, the sensitivity of the data involved and the potential impact. Higher-risk activities, such as exporting confidential records or modifying production systems, shou...
more AI Verified source (Oct 22, 2026)

added 10d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Robin Qiu
Penn State Professor of Information Science and editor-in-chief of Digital Transformation and Society.

I am this person! Add a sourced quote Report comment

votes For and says:

Human-in-Command (HIC) Governance: Complying with international regulatory demands for meaningful human monitoring, the strategic zenith of AI governance requires humans to act as architects rather than tactical micromanagers. Under the HIC model, op...
more AI Verified source (Oct 20, 2026)

added 12d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Sumit Johar
Chief Information Officer at BlackLine; enterprise cloud, cybersecurity, automation, and AI executive.

I am this person! Add a sourced quote Report comment

votes For and says:

AI should not independently handle decisions where accuracy is essential in heavily regulated fields such as finance.
AI Verified source (Oct 19, 2026)

added 13d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
iTechGuides Team
Editorial team at iTechGuides, a technology software directory and guide publisher.

I am this person! Add a sourced quote Report comment

votes For and says:

Require fresh human approval before an irreversible or high-impact action, then log the identity, request, policy decision, approval, and resulting change in a system the agent cannot alter.
AI Verified source (Oct 18, 2026)

added 14d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Jaydeep Singh
General Manager for India at Kaspersky, with over 25 years of enterprise technology and cybersecurity experience.

I am this person! Add a sourced quote Report comment

votes For and says:

Every AI agent should have its own identity and access to only what it needs. [...] High-impact or irreversible actions should require human approval.
AI Verified source (Oct 17, 2026)

added 15d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
John M. Bremen
Managing Director, Board Advisor, and Chief Innovation & Acceleration Officer at WTW; contributor on AI governance and board oversight.

I am this person! Add a sourced quote Report comment

votes For and says:

As agentic AI moves from generating content to making decisions and taking action, boards and C-suite leaders are looking beyond “human in the loop” as their AI oversight model. [...] The newer models don't eliminate the loop. It still belongs in hig...
more AI Verified source (Oct 16, 2026)

added 16d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Kapil Tiwari
Associate Director and Product Management Lead for Hybrid Cloud Services at HCLTech, writing on AI-enabled cyber resilience and governed autonomy.

I am this person! Add a sourced quote Report comment

votes For and says:

The future model is policy-governed autonomy: AI detects, prioritizes and executes approved low-risk actions at machine speed, while accountable leaders retain control of decisions with material operational, legal, regulatory or reputational conseque...
more AI Verified source (Oct 16, 2026)

added 16d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Satya Nadella
Microsoft CEO

I am this person! Add a sourced quote Report comment

votes For and says:

That same kind of mechanistic understanding eludes us in today’s Super Intelligence systems, even as the frontier models powering these systems are now more capable than traditional software systems. We can’t attribute model behaviors and outputs to ...
more AI Verified source (Oct 12, 2026)

added 20d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Bernard Montel
EMEA Field CTO at Tenable and cybersecurity strategist with over 20 years of industry experience.

I am this person! Add a sourced quote Report comment

votes For and says:

The trick is moving away from constant human micromanagement and adopting risk-proportional oversight instead. Low-risk tasks like summarising internal documents can run fully automated, while medium-risk tasks like updating database records can exec...
more AI Verified source (Oct 11, 2026)

added 21d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Mahmoud Ramin
Research director at Info-Tech Research Group with a PhD in Neuroscience, specializing in enterprise IT, AI, and responsible operational deployment.

I am this person! Add a sourced quote Report comment

votes For and says:

As organizations unleash the power of agent autonomy, it raises the pivotal and non-negotiable requirement of enforcing governance. [...] With basic AI systems, risk is much more contained compared to that of an autonomous agent that has access to mu...
more AI Verified source (Oct 9, 2026)

added 23d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Meagen Eisenberg
Chief Marketing Officer at Samsara; enterprise technology executive with experience leading marketing at MongoDB, Navan, and Lacework.

I am this person! Add a sourced quote Report comment

votes For and says:

Governance must be designed into the deployment as well. Organizations should define which actions systems may take autonomously, when human review is required and how decisions can be understood or challenged. The level of oversight should reflect t...
more AI Verified source (Oct 8, 2026)

added 21d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Dipankar Sarkar
Writer on AI data rights and data-monetization policy; author for the Human Data Rights Coalition.

I am this person! Add a sourced quote Report comment

votes For and says:

An agent is a set of actions: read a ticket, look up a customer, draft a reply, update a field, send an email, issue a refund. Each action carries different risk. The right question is: What action needs what kind of human approval? [...] Low reversi...
more AI Verified source (Oct 8, 2026)

added 25d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Tan Kiat How
Singapore Senior Minister of State for Digital Development and Information.

I am this person! Add a sourced quote Report comment

votes For and says:

Existing guidelines and regulations, including risk management processes and oversight, also apply to agentic AI systems. This includes the principle of maintaining human accountability, implementing sufficient controls and guardrails, as well as eff...
more AI Verified source (Oct 7, 2026)

added 25d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Ministry of Digital Development and Information (Singapore)
Singapore government ministry responsible for national digital development and information policy, including AI governance.

I am this person! Add a sourced quote Report comment

votes For and says:

For essential services and other higher-risk applications, stronger safeguards are necessary. [...] We cannot rely on simply telling AI agents what to do. We must also deliberately limit what they can access and do, conduct appropriate test suites, r...
more AI Verified source (Oct 6, 2026) 1 of 2

added 21d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Tamaghn Kasibhatta
AI Product Manager at NxtGen Cloud Technologies, working on enterprise AI infrastructure and governance.

I am this person! Add a sourced quote Report comment

votes For and says:

An agent that can reconcile invoices, route support tickets or prepare a management report may save time with less risk. The same agent should not be allowed to approve a large payment, change a production system, make a hiring decision or send an ex...
more AI Verified source (Oct 6, 2026)

added 26d ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Bryan Harris
Executive Vice President and Chief Technology Officer at SAS, leading its data, analytics, and AI research and development organization.

I am this person! Add a sourced quote Report comment

votes For and says:

State-of-the-art AI agents can have error rates exceeding 25 per cent on complex tasks, a situation that is unacceptable in high-stakes decision-making. Organisations seeking accuracy and repeatability must embed domain expertise into AI-agent workfl...
more AI Verified source (Oct 2, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Julio Rivera
Cybersecurity researcher, business and political strategist, and technology columnist; founder of ItFunk.org.

I am this person! Add a sourced quote Report comment

votes For and says:

Companies do not need to abandon AI. They need to deploy it with discipline. Every agent should have a distinct identity, clearly defined access rights, and an audit trail showing what it did. Where practical, credentials should be short-lived, and p...
more AI Verified source (Oct 1, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Gleb Tsipursky
Behavioral scientist and author

I am this person! Add a sourced quote Report comment

votes For and says:

An agent with production access, payment authority, persistent credentials, or the ability to direct other agents should face stronger evaluation, logging, approval, and incident-reporting expectations. [...] For serious incidents, a report should an...
more AI Verified source (Sep 24, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Alan Turnley-Jones
CEO of NTT Data Middle East and Africa, writing on AI governance, autonomy, and accountable deployment.

I am this person! Add a sourced quote Report comment

votes For and says:

As AI becomes more autonomous the real question for business leaders is where human judgment matters most. What should machines be allowed to do independently? What should trigger human intervention? And who remains accountable for the outcome? [...]...
more AI Verified source (Sep 22, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Fareed Zakaria
CNN host, Washington Post columnist, author, and political commentator.

I am this person! Add a sourced quote Report comment

votes For and says:

This is a systemic problem that we cannot leave to the good graces of private companies. When thinking about regulations, we should focus centrally on how much autonomy we give these systems. A chatbot that answers a question poses one set of risks. ...
more AI Verified source (Sep 18, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Alastair Denniston
NHS consultant and Chair of the National Commission into the Regulation of AI in Healthcare.

I am this person! Add a sourced quote Report comment

votes For and says:

One message came through clearly: people are open to AI improving their care, but only if it is safe, overseen by humans, and if they know when it is being used. These recommendations are designed to make that trust possible, so that the UK can lead ...
more AI Verified source (Sep 10, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Steven Mills
Chief AI Ethics Officer and Managing Director & Partner at BCG

I am this person! Add a sourced quote Report comment

votes For and says:

Deeper human review is concentrated on the AI products that carry the greatest risk.
AI Verified source (Sep 10, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Menlo Security
Cybersecurity company providing browser and AI-agent security products for enterprises.

I am this person! Add a sourced quote Report comment

votes For and says:

At that point, speed matters less than judgment (aka security). A person needs to understand the context, weigh the risk, and decide whether the action should proceed. As agents take on more routine execution, human involvement shifts toward these le...
more AI Verified source (Sep 8, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Australian Government Department of Finance

I am this person! Add a sourced quote Report comment

votes For and says:

Governments should provide clear, simple explanations for how an AI system reaches an outcome. This includes: [...] the implementation of human oversight. [...] When an AI system influences or is used as part of administrative decision making, decisi...
more AI Verified source (Sep 8, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate
Mohan Veloo
Chief Technology Officer for Asia Pacific, China and Japan at F5, specializing in enterprise technology and security.

I am this person! Add a sourced quote Report comment

votes For and says:

Human approval should remain mandatory for actions that could jeopardize human safety, create material financial loss, change production systems, delete or disclose sensitive data, alter security policies, create legal commitments, affect employment ...
more AI Verified source (Sep 7, 2026)

added 1mo ago

Choose a list of delegates

to vote as the majority of them.

Unless you vote directly.

Delegate