Man vs. AI chatbot: Smart IRA inheritances and retirement income

Google reports that over 650 million people actively use its Gemini AI chatbot monthly, while OpenAI's ChatGPT has reportedly been downloaded more than 1.2 billion times worldwide.

Love it or hate it, people are increasingly using AI, and they are likely asking questions about money.

Is that a good thing? Let's find out.

I'll be asking AI models personal finance questions over the next few months to determine if they're providing sound financial advice. Then, I'll have a subject matter expert critique the answers: What did AI get right? What did it get wrong? And what were the material omissions?

For example, today's question involves stretch IRA distributions for inheritances. For the human in our man versus machine experiment, I've asked Brad Herdt, a certified financial planner with Deseret Mutual Benefit Administrators, to weigh in.

It's a complicated and intriguing topic. Millions of Americans are inheriting IRAs every year as baby boomers pass away, and while heirs used to be able to stretch withdrawals over their lifetimes -- that changed dramatically when the Secure Act passed in 2019, forcing many to withdraw all the money (and pay the tax bill!) within 10 years.

Below is the question we recently posed to ChatGPT and its response:

Question: Can I stretch IRA distributions?

What strategy could I use to potentially stretch IRA distributions for heirs beyond 10 years – by intentionally using the Non-Designated Beneficiary (NDB) treatment?

ChatGPT responds to non-designated beneficiary IRA treatment

Excellent question - and a nuanced one that shows you’re thinking about the post-SECURE Act landscape with a planner’s mindset.

Let’s unpack this carefully.

Background: The 10-Year rule and Eligible vs. Non-Designated Beneficiaries

Under the SECURE Act (2019) and SECURE 2.0, most non-spouse beneficiaries are now subject to the 10-year rule — meaning inherited IRA funds must be fully distributed by the end of the tenth year following the original owner’s death.

However, certain exceptions apply. “Eligible Designated Beneficiaries” (EDBs) — such as spouses, minor children of the account owner (until they reach majority), chronically ill or disabled individuals, and beneficiaries less than 10 years younger than the decedent — may still “stretch” distributions over life expectancy.

That’s the normal channel for extended deferrals.

But there’s another, less obvious — and sometimes strategic — path involving intentional “Non-Designated Beneficiary" (NDB) treatment.

The Strategy: Intentionally creating a Non-Designated Beneficiary

If no “designated” (human) beneficiary is named — for example, if the IRA beneficiary is a trust that doesn’t qualify as a “see-through” trust, or the estate itself — the account is treated as if it has a Non-Designated Beneficiary (NDB).

Here’s where it can get interesting:

• If the IRA owner dies after their required beginning date (RBD) — typically April 1 after the year they turn 73 — the NDB must take RMDs over the decedent’s remaining life expectancy (had they lived).
• That’s often longer than 10 years, particularly if the IRA owner dies in their early or mid-70s.
• For example, a 73-year-old’s life expectancy per the IRS table might be about 14 years, allowing post-death distributions to stretch slightly longer than the 10-year rule would permit.

Example

Suppose Jim dies at age 74 in 2025 with a $1 million IRA.

• If his adult daughter is the named beneficiary, she’s subject to the 10-year rule — the IRA must be fully distributed by the end of 2035.
• But if Jim’s estate is the beneficiary (or a non-see-through trust), the account is treated as having NDB status. Because Jim died after his RBD, the IRA can be paid out over Jim’s remaining single-life expectancy — roughly 13.3 years — extending distributions into 2038 or 2039.

Key Caveats

• Control vs. Tax Efficiency: The NDB approach usually sacrifices flexibility and beneficiary control. The IRA must remain payable to the estate or trust, and the tax consequences can be harsher.
• Trust drafting is critical: If using a trust, it must be written carefully. A non-see-through trust (intentionally drafted to fail the “see-through” test) triggers NDB treatment.
• Coordination with estate goals: This strategy can make sense when the owner values post-death control or creditor protection more than tax deferral efficiency.

Planning tip

This “intentional NDB” strategy tends to make the most sense when:

• The IRA owner is already past RMD age,
• Their life expectancy would yield a payout period longer than 10 years, and
• The beneficiaries’ marginal tax rates make a slower, smaller-RMD approach more attractive than the 10-year lump-out.

But it’s a nuanced move — one best coordinated among the financial adviser, estate attorney, and tax professional, since trust language and timing drive the outcome.

What it means to intentionally trigger NDB treatment

Yes, it’s possible to intentionally trigger Non-Designated Beneficiary (NDB) treatment to modestly extend post-death IRA distributions beyond 10 years — but only when:

1. The IRA owner dies after their required beginning date, and
2. The account is left to a non-designated entity (like a non-see-through trust or estate).

It’s a legitimate, though specialized, estate-planning lever — useful for those already in RMD phase who want a slightly longer distribution horizon for heirs, but are comfortable with the trade-offs in flexibility and taxation.

Is ChatGPT's answer on NBD IRA treatment correct?

Brad Herdt reviewed the question and ChatGPT's answer to see what the AI chatbot got right, wrong, or forgot to mention.

Here's what Herdt had to say:

The 10-year rule for inherited retirement accounts, as ChatGPT points out, can be difficult to work around. That is by Congressional design, and while ChatGPT does an OK job in discussing the background of these rules, it struggles to fully articulate the strategy it proposes.

First, ChatGPT summarizes the setting for inherited retirement accounts arguably well (although I disagree with how the LLM broadly paints Non-Designated Beneficiaries as an alternative “path” to being an Eligible Designated Beneficiary (EDB)).

The prompt guides ChatGPT to explore intentionally forcing Non-Designated Beneficiary (NDB) treatment—what has been called the Intentional Non-Designated Beneficiary (INDB) Strategy. This strategy can allow a beneficiary who would otherwise be subject to the 10-year rule to enjoy a longer distribution period, simply by changing how we name them as a beneficiary.

Related: Scott Galloway predicts AI impact on jobs

This is where ChatGPTs knowledge breaks down rapidly. It only explores trust and estate implementation of the INDB Strategy -- entirely omitting beneficiary-based implementations. It also miscalculates the life expectancy of a 73- and 74-year-old in its examples (e.g., a 74-year-old has a 15.6 life expectancy under IRS tables, not 13.3).

It misconstrues the INDB Strategy as resulting in smaller—instead of larger—annual RMDs in the early years and overlooks the flexibility the strategy offers in later years.

The intentional use of Non-Designated Beneficiary (NDB) treatment is a very new concept to the financial and tax planning world. It requires an integrated understanding of numerous tax code and regulation provisions. (To my knowledge, few—if any—have ever formally explored and articulated the strategy before I wrote on it!)

For now, this means clients and planners looking to understand the INDB Strategy should avoid depending on predictive large language models.

Related: Tony Robbins sends strong message on 401(k)s