Boards and executives do not fail at AI because they lack tools. They fail because accountability erodes as decisions accelerate.

Artificial intelligence already shapes hiring, pricing, compliance, product decisions, and risk exposure across organizations.

The challenge for leaders, boards, and founders is not whether to adopt AI.

It is whether the systems surrounding it support accountable human judgment once decisions accelerate, scale, or become opaque.

I work with organizations to evaluate, govern, and oversee AI-enabled systems so responsibility remains visible, decisions remain defensible, and performance remains steady under pressure.

This work is built at the convergence of law, leadership behavior, and governance.

I train frontier AI models, have governed AI risk from inside organizations building and deploying these systems, and have written on AI governance and legislation. That is the vantage point this practice is built on.

How AI risk actually shows up 

• Vendor contracts that obscure responsibility for outputs, bias, or downstream use

• AI tools embedded into workflows without clear decision ownership

• Overreliance on vendor assurances about safety, accuracy, or compliance

• Governance frameworks that exist separately from procurement, legal, and operations

• Leaders expected to oversee systems they were never trained to interrogate

I coined optical ethics by design™ to describe this condition. Accountability thins once AI-managed decisions move faster than human judgment. This is where most organizations believe they have oversight and discover too late that they do not.

Effective AI governance does not live in slogans, policy decks, or one-time assessments.

It lives inside contracts, workflows, escalation paths, and human override mechanisms that still function when pressure is real.

Who this work is for

This work is designed for leaders who retain responsibility for outcomes.

You do not need to know what AI governance means yet. You need to be the person in the room when it matters.

• Board members and advisory boards

• Founders and executive leadership teams

• General counsel, compliance, and risk leaders

• Investors, accelerators, and mission-driven organizations deploying AI

• Organizations operating in regulated or high-stakes environments

• Leaders whose organizations have adopted AI tools without defined decision ownership or accountability for outputs

• Heads of AI governance, responsible AI, and technology risk practices

If AI touches decisions that affect people, trust, safety, or reputation, governance belongs at the center. 

For organizations that also need compliance program assessment, whistleblower program audits, or board preparation on compliance obligations alongside AI governance work, that advisory is available through a dedicated engagement. See Compliance Advisory.

You do not need to have a defined AI governance problem to need this work. You need to be responsible for an organization that uses AI, or is about to.

AI governance becomes urgent through ordinary moments: a board member asks a question nobody in the room can answer, a vendor contract arrives with terms legal has not seen before, or someone starts using an AI tool and nobody knows who owns the output. Those moments do not announce themselves. They accumulate.

When an AI initiative stalls or fails, it is rarely the technology. It is that nobody mapped the stakeholders, nobody established who controls the data, and nobody defined who owns the decisions the system is making.

If your organization uses AI tools, works with AI-enabled vendors, or makes decisions that affect people through automated systems: the question is not whether you have adopted AI. It is whether the people responsible can name those decisions, own them, and account for them when asked.

Decisions I help leaders and boards navigate

• Should this AI system be deployed, paused, or redesigned

• Whether vendor assurances are sufficient or incomplete

• Where human override must be preserved

• How to govern AI across decentralized teams and workflows

This work is about improving judgment before mistakes harden into exposure.

For founders and early-stage teams

AI risk often enters earlier than expected.

Founders and startups face unique challenges:

• Early vendor lock-in through poorly understood terms

• Over-trusting demos, benchmarks, or generic compliance claims

• Building speed without decision discipline

• Scaling tools before accountability is defined

I help early-stage teams build lightweight governance and vendor discipline that protects momentum without importing enterprise bureaucracy.