If you've been wondering whether you're behind on AI, you're not. Plenty of good lawyers haven't touched it. Many more have only tried a free version. And in plenty of firms, one lawyer is quietly using it while the partners down the hall aren't.
Here is what Claude and ChatGPT can do for transactional and litigation work, and where each can hurt you.
If you haven't tried AI yet
Waiting is a reasonable position. Nothing requires you to adopt this tomorrow, and the ethics rules ask you to understand a tool before using it, not to use it by a deadline.
When you're ready, start small with low-risk tasks: rewording a stiff client email, summarizing a public document, or outlining a memo. Leave out client names and confidential facts. Treat the tool like a sharp first-year associate who is fast, eager, and sometimes confidently wrong.
If you're testing free versions
Free accounts are a good way to learn how these tools behave. Keep client information out entirely, because consumer versions may use your conversations to improve their models unless you change the setting.
Try both tools on the same task and compare:
- Transactional work: Claude's drafts tend to read tighter and more formal. ChatGPT tends to be chattier and handy for structured tasks like checklists and rule-based redlines. These are tendencies, not guarantees, and both tools change often.
- Litigation work: Both can invent case citations that look real. ChatGPT's web search can help surface real sources. Claude does well analyzing documents you give it. Either way, every cite must be pulled and read before it goes near a filing.
If you're paying for Claude Pro or ChatGPT Plus
A paid plan buys you more capability and usage. It does not automatically buy you confidentiality. Claude Pro and ChatGPT Plus are personal, consumer plans, and both can use your conversations to improve their models unless you opt out in settings. The opt-out matters, and it's worth confirming the current setting yourself.
If a lawyer in your firm is already using one of these, you don't need to panic. But you do need to know:
- Is the training setting off?
- Is any client information going into it?
- Who reviews the output before it leaves the firm?
Business, enterprise, and API tiers generally don't train on your inputs by default, and they're the better home for client work.
This is also a supervision issue. In a small firm, one lawyer's AI habit becomes every partner's responsibility. A short written policy protects everyone, including the enthusiast.
The Louisiana angle
Louisiana hasn't created a special AI rulebook. Here is what I could confirm:
- The Louisiana Supreme Court issued a letter on AI on January 22, 2024, signaling it is watching the technology's impact on the profession. It doesn't impose specific new duties.
- Existing rules already apply. As reported by Louisiana Legal Ethics, the Court's general counsel concluded that current rules on competence, candor, supervision, and responsibility for work product already cover AI without amendment.
- No AI-specific ethics opinion or statewide mandatory disclosure rule for AI-assisted filings exists, based on what I found. Individual judges may still have their own expectations.
- Act 250 of 2025 addresses AI-generated evidence. It doesn't govern drafting.
That doesn't mean it's risk-free. Louisiana judges have fined attorneys this year for hallucinated authority, including in the Eastern, Western, and Bankruptcy courts, and one federal judge wrote that "ignorance of the risks of AI usage is no longer an excuse."
The bottom line
Claude and ChatGPT are both useful, and neither replaces your judgment. For many small firms, the right first step is not choosing a winner. It's learning on the free versions without client data, then moving to the right plan with clear rules.
On Point Legal Tech helps small and mid-sized firms across the Gulf South get started with AI safely, whether you're just curious or already paying for a plan. Reach out and we'll help you sort out what fits.
Neither tool replaces your judgment. For most small firms, the smart first step isn't picking a winner. It's setting simple rules that let everyone in the firm learn safely.
On Point Legal Tech helps small and mid-sized firms across the Gulf South get started with AI safely. Reach out and we'll help you sort out what fits.