AI Won’t Take the Blame: Business Owners Carry the Risk.

In May 2026, the Chief Justice of New South Wales, Andrew Bell AC, delivered a landmark address; the Harold Ford Memorial Lecture. The lecture focused on “Corporate responsibility and directors’ duties in the era of Artificial Intelligence.” The lecture did not offer prescriptive rules. Instead, it posed a series of sharp questions that cut to the heart of corporate governance in an AI-driven world. Primary among those questions is are you managing your AI Risk?

For business leaders, the message was unmistakable: AI is not just a technology issue; it’s a governance, legal, and personal liability issue for business owners that needs to be taken seriously.

AI Has Already Entered the Boardroom

The Chief Justice framed AI not as a future issue, but as a present and accelerating reality, describing it as a “technological tsunami” with no sign of slowing.

Since the emergence of generative AI tools in late 2022, organisations have rapidly integrated them into decision-making processes, analytics, and operational workflows.

Business takeaway

• Assume AI is already influencing decisions somewhere in your organisation—even if informally.
• Shadow usage (e.g., staff using ChatGPT or Copilot without governance) is a real risk.
• AI risk management must move from IT to board-level oversight.

Directors’ Duties Aren’t Removed, They Are Made Bigger

A central theme of the lecture is that directors’ responsibilities remain personal and non-delegable, regardless of technology.

Even if AI contributes to or generates insights:

• Directors must still exercise independent judgment
• Accountability cannot be shifted to an algorithm

The Chief Justice emphasised that traditional expectations—critical thinking, informed judgment, and oversight—are being strained, not replaced, by AI systems.

Business takeaway

• “The system said so” is unlikely to be defensible in court.
• AI does not dilute liability—it may actually raise the bar.
• Boards must ensure decisions are explainable in human terms, not just machine outputs.

Safe Harbour Protections May Not Apply

One of the most consequential insights concerns existing legal protections for directors, which may not function as expected in an AI context. Traditional safe harbour protections may simply not apply where AI is relied upon for advice. Under Australian corporate law, directors often rely on:

Reasonable reliance on experts (s189)
Delegation protections (s190)
The business judgment rule (s180(2))

The Chief Justice suggested that AI complicates or undermines each of these:

AI is not a “person” or recognised expert
Its outputs may not qualify for statutory reliance
“Black box” decision-making undermines the requirement for independent assessment

Business takeaway

• You cannot assume existing compliance frameworks cover AI usage.
• Legal protections may fail if:
  • You can’t explain how the AI reached its conclusion
  • You didn’t critically interrogate its output

The “Black Box” Problem Is a Legal Risk

AI systems, large language models in particular, often lack transparency. This creates what the Chief Justice highlighted as a major legal challenge:

How can a director make an “independent assessment” of advice if they cannot understand how it was produced? This issue strikes at the core of governance; Business owners are required to show explainability, auditability and evidence when legal questions arise, and the inability to do that when relying on AI puts owners at risk.

Business takeaway

• Treat opaque AI outputs as high-risk inputs, not decision-ready conclusions.
• Require:
  • Documentation of prompts
  • Source validation
  • Human reasoning layered on top
• If challenged, you may need to prove your reasoning, not the AI’s.

AI May Change What “Good Governance” Looks Like

The lecture raises the possibility that AI literacy itself may become a governance requirement.

Key questions posed include:

• Should AI capability be part of the board skills matrix?
• How deeply must directors understand the tools they rely on?
• What documentation should exist around AI-assisted decisions?

There is also concern that AI could:

• Dilute boardroom debate
• Reduce constructive disagreement
• Promote over-reliance on apparently authoritative outputs

Business takeaway

• Business owners should formally assess:
  • AI competence
  • Governance frameworks
• Encourage dissent and challenge, especially when AI outputs appear “confident”.
• Avoid building a culture where AI becomes the default authority in decision making.

Over-Reliance Is a Real and Immediate Risk

The Chief Justice cautioned that the key issue is no longer whether directors will use AI but how far they rely on it.

Over-reliance can:

• Weaken decision-making quality
• Reduce accountability
• Create systemic blind spots

Business takeaway

• Define clear boundaries for AI usage:
• Advisory vs decision-making roles
• Require human validation at critical decision points.
• Stress-test decisions against AI outputs—not just with them.

Litigation Is Inevitable

Perhaps the most sobering takeaway: The Chief Justice hinted that many of these unresolved questions will ultimately be answered through litigation, not policy.

He also noted a broader reality:

• Regulatory lag is inevitable
• But business owners cannot wait for clarity before acting.

Business takeaway

• Courts will likely define acceptable AI use after something goes wrong
• Early adopters without governance will be:
  • Test cases
  • Or cautionary tales

The Strategic Implication: AI Governance Is Now a Board Issue

The lecture ultimately reframes AI from a simple productivity tool to a fiduciary risk category. Business owners already navigating cyber security, financial reporting and workplace safety now need to include AI governance into their thinking in order to protect their business.

Practical checklist for businesses

Executives should immediately ensure:

Governance

• AI usage policy approved at board level
• Clear accountability for AI-driven decisions

Controls

• Documentation of AI inputs/outputs in key decisions
• Human review requirements

Capability

• Board-level AI literacy training
• Defined internal expertise

Risk Assessments

• Customer-facing outputs
• Assessment of where AI influences:
• Financial decisions
• Legal advice

The Final Word

The Chief Justice’s lecture is less about technology than responsibility under uncertainty. The core message for business is clear; You can adopt AI, but you cannot outsource judgment, accountability, or legal liability to it.

Organisations that treat AI as just another tool will fall behind legally. Those that treat it as a governance transformation challenge will be far better positioned when scrutiny inevitably arrives.