AI Governance for HR: The Guardrails That Let You Move Faster

Aarti Mudaliar

Head of Talent , ex- Novartis | NYU Alum

Uvika Sharma

Founder, INTLDA | AI Governance Advisor

You didn't build the AI. You're still liable for it. Here's the proof you need.

You didn't build the AI. You're still liable for it.

McDonald's exposed 64 million applicant records. Samsung lost 270,000 customer records through a vendor's unrotated credentials. Air Canada was held liable for what its own chatbot promised a customer. None of these were AI failures. They were governance failures, and the company answered for it, not the vendor.

Under the EU AI Act, NYC Local Law 144, and Colorado's new AI law, you're the deployer of record for every HR tool your team runs, whether you built it or bought it. Most HR leaders confuse responsible AI with AI governance. Responsible AI is the standard: fair, explainable, accountable. Governance is the proof: who signs off, what gets logged, what you hand a regulator when a decision is challenged.

This course builds that proof live, on tools your team already uses. You leave with three real artifacts: a risk-tiered inventory, a one-page governance map, a vendor accountability assessment. Built by you, in the room, led by two practitioners with 38 combined years in HR operations and AI governance. Not a certificate. Not a template. A file you can hand your CHRO or General Counsel on Monday.

What you’ll learn

Govern the HR AI tools you already use. Know what's high-risk, build a governance map, and question vendors before you sign anything.

Not based on what your vendor calls the tool, based on what it actually decides about people
Apply a three-question risk-tiering logic live to tools your team uses today
Leave with a completed AI inventory, a real artifact, not a case study worksheet

Most HR leaders use these terms interchangeably. They are not the same thing
Responsible AI is the standard. AI Governance is the proof you meet it
The gap between them is where legal and reputational exposure lives

Who approved it, who can override a single output, what gets logged
The document you hand your CHRO, General Counsel, or a regulator if a decision gets challenged
Completed in the session, not taken home as homework

Questions vendors are not expecting, with the answer key: what credible looks like versus what confident-sounding-but-empty sounds
Learn what to accept with conditions, what to escalate, and what should end the conversation
Walk out with a vendor script you can use before the next demo or contract renewal

EEOC has already enforced against AI hiring bias. NYC requires bias audits. The EU AI Act treats people-management AI as high-risk
Know which regulations apply to your organization by location
Leave knowing what "the algorithm decided" means for your personal legal exposure as an HR leader

A risk-tiered inventory entry, a governance map, and a vendor accountability assessment
Together these are the beginning of a real governance file, not a certificate and a folder of notes
Built in four hours on a tool your team uses, not a hypothetical