When an advanced sales producer encounters a unique situation or unfamiliar strategy, AI can provide a useful starting point by supplying background information and helping frame the issue more efficiently than a series of broad searches. The risk arises when that initial information is treated as completed research.
No matter how authoritative an answer appears, it should be confirmed before it becomes part of a client conversation or planning recommendation. Producers remain responsible for any guidance they provide, regardless of how the information was found. Using AI effectively requires understanding what it can do well, how to reduce the risks, and when to turn to a vetted advanced sales resource.
The Risk of an Almost-Right Answer
Most people associate AI errors with “hallucinations,” such as invented facts or citations. For example, in Mata v. Avianca, Inc., in 2023, the court sanctioned attorneys after they filed a brief citing nonexistent cases generated by ChatGPT and continued to stand by the cited cases even after the problem was identified. While AI tools have improved considerably since then, errors have not disappeared.
A more challenging risk may be an answer that is partly correct but materially incomplete. The National Institute of Standards and Technology calls confidently presented false information “confabulation.” A source may be real but fail to support the conclusion, or a general rule may be accurate while an exception, recent change, or jurisdictional limitation is missed. Asking the same system to review its own work is not independent verification and may simply reinforce the same flawed premise.
AI can also be right for the wrong scenario. It only knows the client-specific facts that are included in the prompt, and even a careful prompt can omit important information. A response could be correct under its assumed facts but still be inapplicable to the client. Producers should make it a habit to identify the assumptions behind a response and consider what additional facts could change the conclusion.
The Answer Has Your Name on It
Clients rely on a producer’s professional judgment, regardless of how the research is done. Once an incorrect or incomplete conclusion is shared in a conversation or recommendation, it carries the producer’s implicit approval. Producers should confirm both the rule and whether it truly applies in the particular situation.
The level of review should match the intended use. Asking AI to brainstorm meeting questions poses less risk than relying on it to provide a transaction-specific tax conclusion. The more consequential the answer, the more important it is to check the underlying authority, confirm the facts, and involve a trusted resource or specialist when appropriate.
Producers should also follow company policies governing approved AI tools and permissible uses. Client names, account information, medical or financial details, proprietary illustrations, and other confidential information should not be included in a prompt unless the tool and its intended use have been expressly approved.
An avoidable error or data-handling lapse could lead to unexpected tax or financial consequences, a client complaint, a compliance issue, or an E&O claim. Even without a formal complaint, it can damage the producer’s credibility.
Start with a Trusted Reference
Before beginning with a broad AI search, determine whether the subject is already addressed by a trusted advanced sales reference tool. For many topics, a professionally developed and regularly updated tool such as PGI’s AMO may provide a faster and more reliable starting point.
AMO can help clarify the relevant considerations and provide a reliable reference point for evaluating what AI finds. While it does not eliminate the need to confirm client-specific assumptions, turning to information written by attorneys and industry professionals—peer reviewed, compliance approved, and regularly updated—can remove many of the risks associated with generative AI. It can also serve as a vetted foundation for evaluating what AI finds and determining when an answer may be incomplete, outdated, or inapplicable.
A Better Use for AI
When a question extends beyond the scope of established reference tools, AI may be useful in compiling sources. Traditional search engines work best when the user already knows the right terminology. AI tools with web-search capabilities can interpret plain-language questions better than many search engines and point users toward potentially relevant statutes, regulations, rulings, cases, agency publications, or search terms.
Rather than asking only for a conclusion, a producer might ask:
• “Which primary authorities address this issue?”
• “Are there exceptions to the general rule?”
• “What additional facts could change the analysis?”
• “Have there been more recent developments or competing interpretations?”
• “What terminology should I use to research this independently?”
The response becomes a list of leads to investigate, not the final answer.
Use the Right Tool for the Task
AI should be treated like any other unverified secondary source. It can help formulate questions, identify possible issues, and compile sources, but its conclusions still require independent review.
A trusted advanced sales reference tool such as AMO serves a different purpose. It provides professionally developed, peer-reviewed, compliance-approved content that producers can use as a reliable starting point. AI can broaden the search, while AMO can ground it in vetted research.
Used together thoughtfully, these tools can leave the producer more time for the work that matters to the client: testing assumptions, comparing alternatives, and involving the right specialists. The benefit goes beyond speed, allowing producers to ask better questions and build recommendations worthy of the producer’s—and the client’s—confidence.
