National Institute for Responsible AI Practice

Now certifying · Real estate

Prove your firm uses AI responsibly.

Independent certification for the professional services firms and licensed practitioners who use AI tools, issued by a body that sells none of them.

See the two credentials → Take the free gap assessment

About an hour, no cost and no account. You finish with a ranked picture of where your governance is weakest.

NIRAP Responsible AI Certified Business trust badge

Who NIRAP serves

Built for the firms that license AI, not the ones that build it.

If your practice buys its AI off the shelf and puts the output in front of clients, this was made for you.

The AI governance conversation is aimed at two audiences. One is the companies training the models. The other is the large enterprise with a chief AI officer, a compliance department and a budget to match. Both already have frameworks, and both were written on the assumption that you own and control the system you are governing.

That describes almost nobody. The overwhelming majority of businesses affected by AI do not build it and do not employ anyone whose job is governing it. They license tools from vendors, a seat at a time, and use them every day in client work. Law firms, brokerages, advisory practices, HR consultancies, accounting firms, property managers.

A licensed tool still puts your name on the output. You did not train the model, you cannot audit it, and your vendor contract almost certainly disclaims responsibility for what it produces. What you can govern is how your office uses it, and that is precisely what NIRAP certifies.

Anyone can sell you a policy template or a consulting engagement. Only an independent body can certify you: proof your governance is real, not just written down.

  • One standard, written for firms that license AI rather than firms that build it
  • Sized for a practice with no compliance department and no AI staff
  • Credentials for the firm and for the individual professional
  • Every submission verified by a person, never self-scored
  • No consultant to hire and no engagement to schedule

Guardrails, not brakes.

The guardrails for progress

NIRAP is not here to talk your firm out of using AI. The gains are real, and the practices that use these tools well will outwork the ones that do not. What creates liability is not using AI. It is using it with nothing written down, nobody accountable, and no record of what was checked. A certified firm is one whose people can move faster with confidence, rather than slower while they wait for permission.

55%

Why now

of employers have no AI policy at all.

Not a weak one. None. Meanwhile a majority of American employees now use AI at work, up from roughly a quarter two years ago. Your people adopted these tools on their own, and in most offices nothing written says which tools are allowed, what client information may go into them, or who checks the output before it reaches a client.

That gap is the exposure. It is the same gap in a law firm, a brokerage, an advisory practice and an HR department. You own what your people produce with AI whether or not you approved the tool, and federal regulators have been explicit that using a third party tool does not move the liability off the firm using it.

Policy figure: CNBC and SurveyMonkey Quarterly AI and Jobs Survey, Q3 2026. Adoption figure: Gallup, July 2026.

Choose your path

Certify the firm, or certify yourself.

Two credentials, two different holders. A practice holds one. A licensed professional holds the other. Most firms end up with both.

Certifying now

Real estate. Both credentials are open today. RAICB for brokerages, RAICP for licensed real estate professionals.

Real estate came first because its regulators and its trade association have both already published AI guidance, which gave the standard something concrete to be measured against.

In development

  • Legal
  • Financial services
  • HR and staffing
  • Consulting and advisory
  • Sales
  • Field services

Each track is built against its own profession's rules, so they open one at a time. Tell us which to build next.

RAICB™ Held by the firm
Certifying now · Real estate brokerages

In plain terms: an outside body examined the written rules your office follows when its people use AI, checked the evidence behind them, and certified that the program is real rather than a template someone downloaded.

  • Responsible AI Certified Business trust badge for your firm
  • Detailed assessment report and a public verification page
  • Held by the business, not by one employee who might leave
  • Reviewed by a NIRAP assessor, never self-scored
  • Certification runs for 12 months
The brokerage program → Book a call
RAICP™ Held by the individual
Certifying now · Licensed real estate professionals

In plain terms: a credential proving you personally know how to use AI in client work without creating a liability problem, tested against the rules of your own profession rather than generic tech ethics.

  • Self paced, roughly four hours, then a competency exam
  • Digital badge and certificate issued within 24 hours of passing
  • Listed in the public NIRAP registry
  • Realtor Edition carries seven hours of Texas CE, course 40089-RECE
  • Renewed annually
$295 Get certified → Not my profession

Why it holds up

NIRAP has no stake in the AI tools it assesses.

A credential is worth exactly what the independence behind it is worth. NIRAP sells no AI software and holds no commercial interest in any vendor, model or platform, and our fee is the same whether your firm clears the standard or does not. Here is exactly what NIRAP is, and exactly what it is not.

What NIRAP is

  • An independent standards and certification body with no financial stake in any AI tool
  • A licensed Texas Real Estate Commission continuing education provider, 11365-CEP
  • Author of V-AGE, a standard of more than 100 AI control practices
  • Staffed by assessors who review every submission before a badge is issued
  • Built on frameworks regulators already reference, including NIST AI Risk Management Framework principles, ISO governance standards and NAR published AI guidance

What NIRAP is not

  • Not an AI vendor. We sell no models, no software and no tooling we then certify
  • Not a self assessment. No firm scores itself and no badge issues on assertion alone
  • Not an automated score. A person reviews the evidence
  • Not paid to pass you. The fee is the same whether your firm clears the standard or does not, and no assessor is rewarded for the outcome
  • Not a brake on adoption. We certify how your firm uses AI, we do not tell it to stop
  • Not a legal opinion, an insurance policy or a warranty against liability

The standard behind the badge

V-AGE examines how you use AI, not the technology itself.

V-AGE, the Versatile Adaptive Governance Engine, does not audit the model your vendor built. It examines each way your office actually uses AI, across four dimensions.

1

Inputs

What information goes in. Approved data sources, personal information removed before it reaches the tool, and users trained before they get access.

2

Outputs

What comes out and where it goes. Checks for invented facts, a human reading the work before it reaches a client, and limits on what can be produced.

3

Controls

Who is accountable. A named owner for every use, a person required in the loop on high risk decisions, and a log you can produce later.

4

Constraints

The hard limits. Spending caps, prohibited subjects, regulatory boundaries, and permissions deciding who can use which feature.

Who wrote the standard

V-AGE was not drafted by academics, and it was not drafted by an AI vendor. It was built by people who have carried this risk inside regulated businesses: cybersecurity practitioners, governance and privacy specialists, and business operators who have actually run firms under scrutiny.

That mix is why the standard asks for evidence a working office can actually produce. A control practice written by someone who has never had to satisfy a regulator tends to be either unenforceable or impossible, and V-AGE was validated against real regulatory scrutiny before it was published.

  • Cybersecurity practitioners
  • Governance and privacy specialists
  • Business operators from regulated industries

On the jargon. A control practice is one specific thing your office does and can prove, such as naming the person who approves a new AI tool. A regulatory module is the set of those practices a particular profession's rules require. Your assessment draws from more than 100 practices and more than 10 modules, so what a law firm completes looks nothing like what a brokerage completes.

How firm certification works

Five steps. No consultants. No binders.

Your assessment runs on V-AGE Assess™. You log in, see exactly which controls apply to your firm, and work at your own pace. Most firms finish the baseline in two to four weeks.

01

Discovery call

Sixty minutes to understand how your firm uses AI and which tier fits. Your assessment is configured from this conversation.

02

A tailored dashboard

You see only the controls that apply to your firm and your use cases. There is no generic checklist to wade through.

03

Work at your own pace

The platform explains each item in plain language and drafts missing documents from your own answers for you to review and adopt.

04

A NIRAP assessor reviews

A person examines what you submitted against the V-AGE standard. This human review is what makes the credential credible.

05

Your badge is issued

Trust badge, detailed assessment report and a public verification page. Certification runs for 12 months.

Where the liability lands

"We are careful" is not a defensible answer.

When a client, a regulator or a carrier asks what governance you have in place for the AI tools your people use every day, recollection is not evidence. The exposure looks different by profession, but it arrives through the same three doors.

Employment and hiring

AI used to screen, rank or evaluate people sits under Title VII and the ADEA. The EEOC has been direct that automating a decision does not move responsibility off the employer.

Professional duty of care

When AI assisted advice, filings or documents cause client harm, the claim lands on the practice that delivered them. Your licensing board asks what your firm did, and whether it was written down.

Client data and confidentiality

Client information typed into a personal AI account leaves your control. Most vendor contracts explicitly disclaim responsibility for harm caused by their tools, and prompt histories are discoverable.

State AI statutes are already on the books in Colorado, Illinois and Texas, licensing boards are drafting guidance, and insurers have begun adding AI governance questions to errors and omissions applications.

Real estate, as a worked example

Fair Housing Act civil penalties per violation, before actual damages, attorney fees and injunctive relief. 24 CFR §180.671, amended June 12, 2025. Every profession has its own equivalent schedule.

$26,262No prior violation
$65,653One prior within five years
$131,308Two or more within seven years

Certify now and you build a documented track record. Wait, and you will be assembling one under pressure, after something has already gone wrong.

Start certification →

In development

Tell us who we should certify next.

Real estate is live. These tracks are in development, and the order we build them in is decided by demand. Registering costs nothing and commits you to nothing.

  • Legal
  • Financial services
  • HR and staffing
  • Consulting and advisory
  • Sales
  • Field services

Associations and professional bodies: we are building a national network bringing responsible AI certification to members. Choose the partnership option and we will contact you directly.

No cost, no commitment. Prefer email? Write to contact@nirap.org.