Regulated industries should prohibit AI from making autonomous decisions where fiduciary duty applies

Delivering truly human advice requires context, experience, patience and empathy. It also requires accountability. Fiduciary advice means someone stands behind every recommendation and is responsible for its impact on a client's life. AI Verified source (2026)

Who's really responsible, and can people really be relying on a product to do this if it's not being backed up by a corporation with a fiduciary duty? AI Verified source (2026)

The answer right now is, clearly, AI has the [financial] expertise. What they don't have is that fiduciary duty. [...] we're not going to get to the point where we can fully delegate these [financial] decisions. AI Verified source (2026)

Artificial intelligence has the potential to expand access, improve efficiency and strengthen retail financial services, but only if trust and accountability remain firmly at the center. The priority now is operational clarity: clear guidance on how ... more Unverified source (2026)

We are moving from AI-as-tool, where a human uses a machine to support their own analysis or recommendations, to AI-as-agent, where the machine initiates, decides, and executes, potentially without human involvement. A fiduciary gap emerges in a worl... more Unverified source (2026)

Companies that provide AI services don't appear to receive compensation for their advice to retail investors, and therefore aren't fiduciaries. [...] It is the advisor — and not the company backing the AI platform — that would be liable. Unverified source (2026)

AI raises the floor. Humans raise the ceiling. Technology can accelerate routine analysis and connect information in powerful ways. But judgment — understanding context, interpreting signals, navigating ambiguity and building trusted relationships — ... more Unverified source (2026)

We don't want it to be fully autonomous and doing things which we have no control over. So we're putting in all the necessary guardrails. Unverified source (2026)

In practice, 'show me your guardrails' will increasingly mean 'show me your workflow.' Organizations will stop prioritizing maximum autonomy and instead demand systems that constrain AI through structured, logged processes, making explainability and ... more Unverified source (2026)

All 'recognized privileges' require 'a trusting human relationship' with 'a licensed professional who owes fiduciary duties and is subject to discipline' — conditions an AI platform cannot meet. Unverified source (2026)

The real value of generative AI in audit isn't just about speed; it's about shifting the focus of the practitioner. [...] We can re-allocate auditors' focus from low-value, repetitive tasks to the high-value areas that truly require their professiona... more Unverified source (2026)

I want to automate the first draft of a lot of content for our users, but we're going to do it in a way that safeguards and preserves human judgment. Unverified source (2026)

EY's human-led, AI-powered audit of the future is a market-leading example of enterprise AI in action, delivering greater value, deeper insight and increased confidence for our clients and stakeholders. Unverified source (2026)

That means clear guardrails on how and where it's used, strong information-security controls, rigorous model validation, human accountability for decisions, and ongoing evaluation as the technology evolves. [...] The human makes the decisions. Unverified source (2026)

AI is great at very directed tasks, it's not quite great at broad deterministic reasoning yet. It can make a recommendation but it's not sure enough to take action. That's where humans come in. Verified source (2026)

Plugging [AI] into an insurer, whilst technically possible, today generates pretty catastrophic consumer outcomes. It can hallucinate on consent, it can hallucinate on terms and conditions, it can hallucinate on prices. Verified source (2026)

Professionals are responsible for outcomes and errors carry real consequences. [...] Almost right isn't good enough, whether in a courtroom, a boardroom or an audit. Verified source (2026)

The technology remains absolutely crap at judgment. [...] It can't do investing right now. AI Unverifiable source (2026)

We require human-in-the-loop for all clinical decision and diagnostic support tools. AI Verified source (2026)

AI does not independently diagnose, order, or treat — it can strongly recommend and prebuild orders, but a clinician must click 'accept.' AI Verified source (2026)

In today's world, I just don't trust the computer to say 'no.' AI Verified source (2026)

A major worry is that wrongful denials may be occurring as a result of a lack of meaningful human review of recommendations made by AI. AI Verified source (2026)

Agentic tools can thus handle the 'how,' while clinicians handle the 'why.' AI Verified source (2026)

Fully autonomous, unsupervised legal decision-making, especially in high-stakes or regulated contexts, will remain both unrealistic and unacceptable. AI Unverifiable source (2026)

More sophisticated AI requires more human oversight, not less. AI Verified source (2026)

We can't let generative AI into the practice of medicine without human oversight. AI Verified source (2026)

Full autonomy may sound efficient, but in legal workflows, it is often inappropriate. Sensitive actions should require human checkpoints and confirmation. AI Verified source (2026)

AI should not replace physician judgment. Physicians are reporting more denials and longer wait times for peer-to-peer reviews, straining clinic resources and slowing access to care. AI Verified source (2026)

I'm skeptical that you can just sort of replace clinical judgment with artificial intelligence, and that's where you need to have the guardrails. AI Verified source (2026)

The principle is that we make better human decisions by getting AI to analyse it for us. AI Verified source (2026)

The Business Judgment Rule will endure only if it rewards directors who exercise not ritual compliance, but authentic discernment—directors who practise what we call true care and algorithmic loyalty lived. AI Verified source (2026)

Tight governance, and the right human oversight, is the difference between breakthroughs and mistakes. AI Verified source (2026)

But as this technology advances, it is critical that augmented intelligence be designed to enhance—not replace—physicians. AI Verified source (2026)

I'm going to fight tooth and nail against the Trump administration's efforts to substitute real health care with poorly-regulated AI and robots. Because our country's seniors deserve real care from real doctors, not a Silicon Valley robot. AI Verified source (2026)

Algorithmic detection of possible misconduct should not and cannot supplant the considered judgment of our commissioners and staff, nor can it serve as the sole basis of an SEC enforcement action. AI Verified source (2026)

We as CFP professionals need to be in a position to utilize these tools but do it in a way that honors our fiduciary responsibilities and the privacy of the families we serve. Unverified source (2025)

Delegating decisions to a machine does not absolve the human fiduciary from oversight. Today, AI cannot fulfill these fiduciary duties without human oversight. Unverified source (2025)

Our staff's expertise and knowledge is and will remain crucial to successfully leveraging GenAI, especially when maintaining a 'human-in-the-loop' as part of our governance and controls. Unverified source (2025)

We're continuing to build out our AI capabilities with increasingly sophisticated agents in KPMG Clara to enable our auditors to more effectively respond to risks and deliver deeper audit insights. As we accelerate our adoption of this innovative tec... more Unverified source (2025)

Leveraging technology that complements and amplifies the professional judgment of our auditors is essential to shaping the future of the profession. Unverified source (2025)

ChatGPT may process data at lightning speed, but it's the CPA and management accountant who contextualizes it, adding insights and judgment, ensuring businesses not only thrive but also maintain their ethics and integrity. Unverified source (2025)

While this black-box condition may be an uncomfortable spot for an engineer, it is a legally precarious position for a fiduciary. Unverified source (2025)

AI has the potential to improve audit quality, but the human element cannot be removed. Unverified source (2025)

Medical decisions must be made by physicians and their patients without interference from unregulated and unsupervised AI technology. AI Verified source (2025)

Healthcare executives should define clear autonomy boundaries tied to clinical risk, invest in engineering controls that enforce policy automatically and hold AI systems accountable through measurable safety indicators. AI Verified source (2026)

We don't want to stand in the way of innovation that generally serves consumers. But we do want to make sure that it is used transparently, fairly and in ways that hold up to scrutiny. AI Verified source (2026)

AI is already shaping financial services, but its longer-term effects may be more far-reaching. This review will consider how emerging uses of AI could influence consumers, markets and firms, looking towards 2030 and beyond. AI Verified source (2026)

What does good look like when you can't always have a human in the loop, or even able to understand why certain trading decisions are taken or strategies pursued? AI Verified source (2026)

Intelligent use [of AI solutions] can, should, and will catalyze a transformation of the technology of investment management. AI Verified source (2026)

With agentic AI, experienced auditors would be supported by autonomous agents, allowing them to dedicate more attention to complex and judgment-oriented procedures. Unverified source (2025)

The government doesn't need to be regulating everything. We need to get out of the way of clinical decision support tools that doctors find useful. AI Verified source (2026)

Health care will become autonomous because the survival of health systems depends upon it. AI Verified source (2026)

Full automation in areas like diabetic retinopathy screening could significantly improve patient care and 'time to therapy.' AI Verified source (2026)

We already have an AI assistant we built for our home blood pressure monitoring program — that right now still has a human in loop for doing the titrations of the meds. Five years from now, we're not going to have a human doing those titrations. AI Verified source (2026)

In these situations, especially if the outcomes are below expectations, there must still be accountability within the asset owner organisation, and the buck ultimately stops, just as it does now, with the CIO. AI Verified source (2026)

At some stage, we're going to trust that the agent can make some of the decisions and we just monitor what it does. AI Verified source (2026)

We're redefining what it means to be an insurance agent — with fully autonomous AI agents that will eliminate human error, offer superior client interactions 24/7, and fundamentally alter industry expectations. Unverified source (2025)