Washington Just Signed an AI Chatbot Law. Oregon and Georgia Are Next. Here's Your Compliance Checklist.

March 31, 20267 min readBy The Crossing Report

On March 24, 2026, Washington Governor Bob Ferguson signed HB 2225 into law.

It's the first chatbot safety law in the Pacific Northwest — and the compliance clock is now running. The effective date is approximately June 22, 2026, 90 days after signing.

If your firm has any client-facing AI tool — an intake chatbot, a FAQ bot, an AI scheduling assistant on your website — you have about 83 days to get compliant.

Get the full picture. Go premium.

Weekly intelligence briefings, deeper analysis, and direct access to the full archive.

Oregon and Georgia are next. Both states passed their chatbot bills in March with near-unanimous votes and are awaiting governor signatures. When they sign, the legal obligations will extend to two more states. The private right of action in Oregon's bill makes non-compliance a direct liability risk — $1,000 per violation, no cap stated.

Here's what professional services firm owners need to know.


What the Laws Actually Require

All three laws (Washington HB 2225, Oregon SB 1546, Georgia SB 540) follow the same basic framework:

1. Disclose that you're AI, not a human. At the start of every interaction, users must be told they are talking to an automated AI system. This applies whether the tool is on your website, embedded in a client portal, or delivered via messaging app.

2. Repeat the disclosure. Washington HB 2225 requires disclosure at least every three hours during an ongoing adult interaction, and every hour when the user is a minor. Oregon and Georgia have similar requirements.

3. No manipulative behavior. The laws prohibit chatbots from simulating human emotional responses in ways designed to encourage continued engagement or dependency. An intake bot that says "I really want to help you through this difficult time" in a way designed to keep users talking isn't a compliant chatbot.

4. Crisis referrals. Any AI assistant that handles conversations where mental health distress or self-harm might come up — which includes legal consultations, financial hardship discussions, and employment disputes — must have crisis hotline referral language embedded. The National Suicide Prevention Lifeline (988) must be accessible.

5. Enhanced protections for minors. If your AI tool can be accessed by users under 18 — including intake tools on websites without age gates — additional protections apply, including parental controls and more frequent disclosure repetition.


Which Professional Services Firms Are In Scope

If your firm is in Washington, Oregon, or Georgia, and you operate any of the following, you are likely a covered chatbot operator:

  • A chat widget on your website powered by any AI tool (Drift, Intercom, Freshdesk, HubSpot AI, or a custom solution)
  • An AI intake tool that collects information from potential clients
  • An AI FAQ assistant that answers questions about your services, fees, or process
  • An AI scheduling tool that interacts conversationally rather than presenting a simple calendar
  • Any AI tool embedded in a client portal that responds to questions

Firms NOT in scope: Purely calendar-booking tools that don't interact conversationally (like Calendly with no chat function) are not chatbots under these laws. Internal tools used only by your staff — not client-facing — are also not covered.


The Oregon Liability Risk Is Different

Washington HB 2225 is enforced by the Washington Attorney General with civil penalties. That's meaningful but requires the AG to bring a case.

Oregon SB 1546 is structurally different: it includes a private right of action. Individual users can sue directly, with statutory damages of $1,000 per violation. No cap is stated in the bill.

This matters for a 10-person firm. A client-facing AI tool that interacts with 50 inquiries per month without the required disclosure language is not a theoretical violation — it's $50,000 in potential monthly exposure if every user filed a claim. Statutory damages under private right of action are exactly the kind of liability that attracts plaintiff's attorneys.

Oregon's law takes effect January 1, 2027 if Governor Kotek signs — which the 52-0 House vote makes likely. You have approximately nine months. That's enough time to get compliant, but not enough time to ignore it.


The Three-State Compliance Template

Here's the practical good news: one compliant disclosure policy covers all three states.

Step 1: Add a disclosure banner at the start of every AI chat interaction.

This should appear as the first message the user sees when they open the chat window, before any other interaction:

"You're chatting with an AI assistant — this is an automated tool, not a member of our team. For questions about your specific matter, please call us at [phone number] or email [address]."

If any part of your clientele is under 18 (family law, estate planning, educational consulting), add: "If you are under 18, please have a parent or guardian assist you."

Step 2: Add periodic disclosure for longer conversations.

If your chat tool supports conversation triggers, configure a message to appear after 60 minutes of continuous interaction:

"Reminder: you're still chatting with an AI assistant. For matters that require legal [or professional] judgment, please contact our team directly."

Step 3: Add crisis referral language.

If your firm handles matters where clients may be in distress — legal disputes, financial hardship, employment termination, family law — add a standing response trigger for distress signals:

"If you're going through something difficult and need support, please reach out to the 988 Suicide and Crisis Lifeline (call or text 988). Our team is here to help with your [legal/financial/professional] matter when you're ready."

Step 4: Verify your tool vendor's compliance posture.

The major AI chat platforms (Intercom, Drift, Freshdesk) are actively updating their products for state chatbot compliance. Check their release notes or contact your vendor to confirm whether they've built Washington HB 2225 compliance into the product defaults. Do not assume they have.


Your 83-Day Compliance Checklist (Before June 22, 2026)

This is the minimum viable checklist for a professional services firm with one client-facing AI tool in Washington, Oregon, or Georgia:

  • Identify every AI-powered client-facing tool your firm uses. Include website chat, scheduling tools with conversational AI, and tools embedded in your client portal.
  • Add disclosure language at the start of every interaction using the template above.
  • Configure timed re-disclosure for sessions exceeding 60 minutes (Washington standard for adult users).
  • Add crisis referral language — 988 hotline minimum, embedded as a standing response to distress signals.
  • Review your vendor's compliance documentation to understand whether their product meets the requirements natively or requires manual configuration.
  • Remove or clearly document any emotionally manipulative AI responses in your chatbot's conversation design.
  • If your site is accessible to minors, add age verification or heightened protections per HB 2225.
  • Log the date you completed compliance changes. In any enforcement action, documentation of a good-faith compliance effort matters.

The Bigger Picture: This Is the Start, Not the End

Washington, Oregon, and Georgia are three of at least five states with signed or near-signed AI chatbot legislation in March 2026 alone. The regulatory pattern has shifted from "bills are moving through legislatures" to "bills are becoming law."

Professional services firms that treat AI chatbot compliance as optional are running a liability risk that gets more expensive each month a new state adds its name to the list.

The firms that will benefit most from this moment are the ones that build a compliant AI client communication posture now — not in response to a complaint or enforcement action, but as a proactive professional standard. In the same way that GDPR-ready firms handled the privacy wave more gracefully than those who waited, the AI compliance window is open right now.

The Washington effective date is June 22. Oregon's governor could sign any day. Georgia's governor is expected to sign soon after.

The compliance conversation isn't coming. It's here.


What to do this week: Identify your firm's client-facing AI tools. If you have one, add the disclosure language in the template above. It takes 15 minutes and covers your Washington, Oregon, and Georgia exposure in a single pass.

Regulatory status note: Oregon SB 1546 and Georgia SB 540 are awaiting governor signatures as of March 31, 2026. This article will be updated when signatures are confirmed.

Frequently Asked Questions

Does Washington HB 2225 apply to professional services firms with AI intake tools?

Yes, if your firm uses a chatbot, AI scheduling assistant, FAQ bot, or any AI-powered tool that interacts with potential or existing clients on your website or through a messaging platform, HB 2225 applies. The law requires chatbot operators — including professional services firms — to disclose that users are interacting with AI, not a human. The effective date is approximately June 22, 2026 (90 days after March 24 signing). For minor users, disclosure is required every hour. For adult users, every three hours during an active interaction.

What does 'disclosure' mean under Washington HB 2225?

Disclosure means clearly informing the user that they are interacting with an automated AI system, not a human. A compliant disclosure might read: 'You are chatting with an AI assistant. This conversation is automated and not a human attorney [or accountant, consultant, etc.]. For questions about your matter, please call our office.' The disclosure must appear at the start of each interaction, and for interactions involving minors, must repeat at least once per hour. The law also prohibits manipulative chatbot behavior — AI that simulates human emotional responses to encourage continued engagement — and requires crisis referral protocols (mental health/suicide hotline information) to be embedded in any AI assistant.

What is Oregon SB 1546 and when does it take effect?

Oregon SB 1546 is the state's consumer-facing AI chatbot safety law. It passed the Senate 26-1 and House 52-0 and as of late March 2026 awaits Governor Tina Kotek's signature. If signed, it takes effect January 1, 2027. The most significant element for professional services firms: SB 1546 includes a private right of action with statutory damages of $1,000 per violation and no stated damages cap. A single AI chat interaction that fails the disclosure requirement could trigger a $1,000 statutory damages claim from a user. For a firm processing 50 intake inquiries per month through an AI tool, a non-compliant interaction pattern is a material liability exposure.

Does Georgia SB 540 create the same liability as Oregon SB 1546?

Georgia SB 540 passed both chambers in late March 2026 and awaits Governor Kemp's signature. It follows a similar structure to Oregon SB 1546 — mandatory disclosure that users are interacting with AI, restrictions on manipulative behavior toward minors, parental control requirements, and crisis referral protocols. The enforcement mechanism is Attorney General civil enforcement with civil penalties, rather than Oregon's private right of action model. For a professional services firm: the practical obligations are nearly identical between the two states. If your AI tool is compliant with Washington HB 2225, adding the required disclosure language covers the Oregon and Georgia requirements simultaneously.

Which professional services firms are most at risk from these laws?

Any firm with a client-facing AI tool in Washington, Oregon, or Georgia is directly in scope. This includes: law firms with AI intake assistants or chatbots that answer initial questions about cases; accounting firms with AI tools that answer client questions about tax filings, appointments, or services; consulting firms with AI assistants on their website or in client portals; staffing agencies using AI to screen job seekers or answer applicant questions; and marketing agencies using AI chat to manage client inquiries. If your website has a chat widget powered by Drift, Intercom, Freshdesk, or a custom AI tool — and that tool uses AI to respond — it is a covered chatbot under these laws.

What about firms operating in multiple states — do they need to check each state separately?

Yes, but the practical compliance posture is the same across Washington, Oregon, and Georgia: disclose that the user is interacting with AI at the start of the conversation, repeat that disclosure periodically, avoid manipulative behavior, and embed crisis referral language. Building one compliant disclosure template that meets Washington HB 2225 (the strictest effective-date requirement) will likely satisfy Oregon and Georgia requirements when those laws take effect. Monitor your state bar's guidance and the Troutman Pepper State AI Law Tracker for updates.

Is the New York RAISE Act relevant to small professional services firms?

No. The New York RAISE Act chapter amendments signed in March 2026 target frontier AI developers — specifically companies with $500 million or more in annual revenue that train and publicly offer generative AI systems. A law firm, accounting firm, consulting firm, staffing agency, or marketing agency does not develop frontier AI models and is not a covered entity under the RAISE Act. The law is directed at OpenAI, Google, Anthropic, and similar large AI providers. Professional services firms are not directly regulated by this law.

Get the weekly briefing

AI adoption intelligence for accounting, law, and consulting firms. Free to start.

Free weekly digest. No spam. Unsubscribe anytime.