AI Readiness: 7 Proven Steps to Prepare Yourself and Protect Your Clients from Costly Mistakes

AI readiness isn’t a buzzword you can skip past anymore. If you work with clients in any advisory capacity, whether you’re a lawyer, accountant, consultant, financial planner, or marketing agency owner, AI tools are already showing up in your client conversations, whether you brought them up or not. Clients are asking if they should use ChatGPT to draft contracts. They’re wondering if AI-generated financial projections are safe to show investors. Some have already used AI in ways that landed them in legal trouble, and they don’t fully understand why.

This puts professionals in an uncomfortable spot. You’re expected to have an informed opinion on tools that didn’t exist five years ago, while also protecting clients from mistakes that are now showing up in courtrooms, regulatory filings, and news headlines. The temptation is to either dismiss AI entirely or embrace it uncritically. Both reactions are risky.

This article walks through what it actually means to prepare yourself for AI, how to support clients without encouraging reckless use, and what we can learn from real cases where people got AI badly wrong, including fabricated court citations and multimillion-dollar deepfake fraud. The goal isn’t to scare you away from AI. It’s to help you use it the way a careful professional should: with verification, boundaries, and a clear sense of where the risks actually are.

What AI Readiness Actually Means for Professionals

AI readiness gets thrown around loosely, so it’s worth being specific. It doesn’t mean knowing how to write a clever prompt. It means having enough working knowledge of how these tools behave, where they fail, and what obligations you still carry as a professional, so that you can use AI without quietly handing off your judgment to a system that doesn’t actually understand your client’s situation.

A professional with real AI readiness can answer three questions without hesitating:

• What is this tool actually good at, and where does it tend to be wrong?
• What happens if the AI-generated output is incorrect and I pass it along anyway?
• What am I still legally and ethically responsible for, regardless of what tool produced the work?

If you can’t answer those three questions for the tools you’re using, you’re not ready to advise clients on them yet, even if you’ve been using AI daily for months.

Why This Matters Now: Real Cases of AI Misuse

It helps to look at what’s already gone wrong, because these aren’t hypothetical warnings. They’re documented incidents involving professionals who should have known better, and in some cases, clients who paid the price for someone else’s shortcut.

Fabricated Legal Citations: The ChatGPT Lawsuit Case

In one of the most widely reported cases of professional AI misuse, a New York attorney used ChatGPT to help research a legal brief in a personal injury case against an airline. The brief included six fictitious case citations generated by the chatbot, and a federal judge ordered the lawyers and their firm to pay a $5,000 fine. The attorney later admitted he used the tool for the first time and was unaware that its content could be entirely fabricated, despite the AI providing realistic case names, captions, and summaries in standard legal format.

What makes this case worth remembering isn’t just the fine. The judge noted there was nothing inherently improper about lawyers using AI for assistance, but said legal ethics rules impose a gatekeeping role on attorneys to verify the accuracy of their filings. In other words, the tool wasn’t the problem. Skipping verification was.

This wasn’t an isolated incident, either. In a separate case in Alabama, three private attorneys representing the state’s prison system were sanctioned and removed from a federal case after submitting filings containing completely made-up case law generated by ChatGPT. One of the attorneys later acknowledged he relied on the AI-generated output without confirming that the citations were valid or applicable. The judge ordered the sanctions shared with every client, judge, and opposing counsel connected to the firm.

Deepfake Fraud: The $25 Million Video Call

Legal research isn’t the only area where AI misuse has caused real financial harm. In early 2024, a finance worker at a multinational company in Hong Kong was tricked into paying out $25 million to fraudsters who used deepfake technology to pose as the company’s chief financial officer during a video conference call. The worker believed he was on a call with several colleagues he recognized, but all of them were deepfake recreations.

Believing everyone on the call was genuine, the employee agreed to transfer roughly $25.6 million across multiple transactions to accounts the fraudsters controlled. The scam was only uncovered afterward, when the employee checked in with the company’s head office. Hong Kong police described it as one of several recent cases where criminals used publicly available footage to build convincing fake identities for financial fraud.

These two cases, one involving sloppy use of AI by a professional, the other involving criminal use of AI against a professional, show the two sides of the same coin. AI readiness means protecting yourself from both kinds of failure: the mistakes you might make, and the attacks you might not see coming.

Building Your Own AI Readiness

Before you can responsibly guide clients, you need your own house in order. Here’s what that actually looks like in practice.

Learn the Tools Before You Recommend Them

Don’t recommend or use an AI tool with clients until you’ve tested it on low-stakes work first. Run it through scenarios close to what you actually handle, and pay attention to where it gets things wrong, not just where it impresses you. Every AI system has failure patterns. Large language models, for instance, are known to produce confident-sounding but false information, a behavior often called “hallucination.” If you don’t know what that looks like in the tool you’re using, you won’t catch it when it shows up in a client deliverable.

Understand the Legal and Ethical Boundaries You Still Operate Under

AI doesn’t change your professional obligations, it just adds a new layer you have to manage on top of them. Lawyers still owe clients competent representation. Financial advisors still owe fiduciary duty. Accountants are still bound by accuracy standards. The American Bar Association and other professional bodies have made clear that AI readiness includes understanding that responsibility for a filing, report, or recommendation doesn’t transfer to the software that helped produce it. If you want a fuller picture of how courts are approaching this, the OECD’s AI incident tracker documents a growing list of professional AI misuse cases worth reviewing.

Build a Verification Habit, Not Just a Workflow

The single clearest lesson from the ChatGPT legal citation cases is that AI output needs to be checked against a source you trust, every time, not just when something seems off. Build this into your process the same way you’d build in a second set of eyes on a major filing:

• Never submit or send an AI-generated fact, citation, or figure without independently confirming it exists and says what the AI claims it says.
• Treat AI output as a draft from a junior researcher who sometimes makes things up confidently. Useful, but not final.
• Keep a record of what was AI-assisted and how it was verified, in case anyone ever asks.

Supporting Clients in Their AI Use

Once your own practices are solid, the next job is helping clients avoid the same mistakes, often before they even realize they’re at risk.

Set Clear Expectations Early

Many clients assume AI tools are more reliable than they are, largely because the output reads so confidently. Part of your job is resetting that expectation. Explain, in plain terms, that AI tools can produce wrong information that looks completely correct, and that this is a known limitation, not a rare glitch.

Help Clients Build an AI Use Policy

If your clients run a business, encourage them to put basic guardrails in writing rather than leaving AI use to individual judgment calls across the team. A simple policy should cover:

• Which AI tools are approved for use, and for what kinds of tasks
• A requirement that AI-generated content involving facts, figures, or legal language be verified by a human before it’s used externally
• Clear rules around what client or company data can and can’t be entered into AI tools
• A reporting process for anything that looks like AI-enabled fraud, including suspicious video calls or voice messages

Teach Verification as a Skill, Not an Afterthought

Clients don’t need a technical understanding of how AI models work. They need a habit of checking before trusting, especially around anything involving money, identity, or legal language. Walk them through a basic rule: if an AI tool produced it, and the stakes are high if it’s wrong, verify it against an independent source before acting on it.

This is especially important given how convincing deepfake and voice-cloning scams have become. The Hong Kong case wasn’t a low-budget attempt. It fooled someone on a live video call with multiple participants. Clients should know that a face and voice they recognize on a screen or phone call is no longer automatic proof of identity, particularly when a request involves an unusual payment or urgent secrecy.

Red Flags That Should Slow You and Your Clients Down

Certain situations call for extra caution before trusting AI-assisted work or an AI-mediated interaction:

• A request for an urgent, unusual, or secret financial transaction, especially if it came through video, voice, or a chat message
• AI-generated content citing sources, cases, studies, or data points you can’t independently locate
• Pressure to skip the usual verification or approval steps because “it’s already been checked”
• Any AI output related to legal, medical, financial, or compliance matters that hasn’t been reviewed by a qualified human
• A caller or video participant pushing you to act before you have time to confirm anything through a separate channel

Practical Steps to Move Forward

Bringing this together, here’s a short, usable checklist for both your own practice and your client conversations:

1. Test any AI tool on low-stakes work before using it on client-facing deliverables.
2. Verify every fact, citation, and figure an AI tool produces, without exception.
3. Keep your professional and ethical obligations front and center, regardless of what tool did the drafting.
4. Help clients write a simple, clear AI use policy if they don’t already have one.
5. Train clients and staff to treat unusual payment requests, even from familiar faces and voices, as something to confirm through a second channel first.
6. Document how AI was used and verified in any work product that matters.
7. Stay current on real cases of AI misuse, since the specific failure patterns keep evolving.

Conclusion

Preparing yourself and your clients for AI isn’t about becoming a technologist or swearing off these tools altogether. It’s about treating AI the way you’d treat any powerful but imperfect assistant: useful for drafting, research, and first passes, but never a substitute for your own verification and professional judgment. The lawyers who were sanctioned for submitting fabricated ChatGPT citations and the finance worker who lost $25 million to a deepfake video call didn’t fail because AI is inherently dangerous. They failed because the usual checks, verifying a citation, confirming a transaction through a second channel, got skipped. Build those checks back into your own workflow, teach your clients to do the same, and AI readiness stops being a vague goal and becomes a habit that actually protects the people relying on you.