PTE 2020-02 Rollover Documentation in a Breakaway: What AI Can Draft and What a Human Must Sign
PTE 2020-02 still governs rollover advice in 2026, even after the 2024 Fiduciary Rule was vacated. It requires a documented best-interest analysis that compares the plan and the IRA on fees, services, investments, and available alternatives, plus written fiduciary and conflict disclosures. AI can draft the comparison and assemble the disclosures, but a licensed advisor must sign the recommendation.
A breakaway is not only an account-transfer problem. Buried inside a book move is a compliance obligation that has nothing to do with ACATS and everything to do with what the advisor recommends: when a departing advisor tells a client to roll an old employer retirement plan into an IRA at the new firm, that recommendation is fiduciary investment advice, and Prohibited Transaction Exemption 2020-02 sets the conditions the advisor and firm must meet to get paid for it. Handled by hand across a book full of clients with 401(k) assets, the documentation is a slow, error-prone grind. This is what the rollover file must actually contain, where AI can do the assembling, and where a licensed human has to own the judgment and the signature.
FastTrackr AI treats this the way it treats the rest of a transition: the machine drafts, gathers, and validates, and a professional decides. AI can build the comparison and pre-fill the disclosures for a hundred rollover files without a human retyping a single fee table, but it does not, and must not, make the best-interest determination.
Why PTE 2020-02 still applies in 2026
There is real confusion here, so start with the status. The Department of Labor's 2024 Fiduciary Rule, which would have broadened who counts as a fiduciary, was effectively undone: the DOL withdrew its defense in late 2025 and federal courts in Texas finalized that outcome in early 2026. But PTE 2020-02 is a separate, older instrument, and it remains in effect. The exemption's core rollover obligations, performing a best-interest analysis before recommending a rollover and giving the client a written explanation of the specific reasons it is in their interest, have not changed. The clearest current-status summary is Groom Law Group's analysis of the DOL's PTE 2020-02 FAQs, and the mechanics of compliance are laid out in Kitces on complying with PTE 2020-02 rollover requirements.
The trigger matters for a breakaway specifically. If the advisor provides ongoing advice for a fee and recommends the rollover, the recommendation is covered and the exemption's conditions apply. Not every account move in a transition is a rollover recommendation, though, and telling the two apart is the first place a transition tool earns its keep.
Which transition events actually trigger the exemption
A book move contains several kinds of account transfer, and only some are rollover recommendations. A taxable brokerage account moving firm to firm via ACATS is not a rollover; it is a re-registration and a transfer. An IRA moving trustee to trustee is generally a transfer, not a distribution, and does not by itself trigger the rollover analysis. But an employer plan, a 401(k) or 403(b), that the advisor recommends moving into an IRA is the event PTE 2020-02 is built around, because the client is leaving a plan and the advisor is recommending they do it.
Getting this sorting right at the start of a transition is not trivial when a departing advisor has hundreds of households, each with a different mix of taxable accounts, existing IRAs, and old employer plans. Misclassify a rollover as a simple transfer and the file is missing its required best-interest documentation. Misclassify a transfer as a rollover and you generate compliance work that was never needed. AI is well suited to the sorting step, reading the incoming account data and flagging which moves are plan-to-IRA rollovers that need the full analysis, while a human confirms the classification on anything ambiguous.
What the rollover file must contain
Once an account is confirmed as a rollover recommendation, the file has to demonstrate a real best-interest analysis and carry the required disclosures. The DOL's guidance, and the practitioner breakdown in Fred Reish's analysis of the factors to evaluate for a rollover recommendation, converge on four comparison points the advisor must evaluate and document:
- The alternatives to a rollover, including leaving the assets in the employer plan if that option is available.
- A comparison of fees and expenses between the current plan and the proposed IRA, noting whether the employer subsidizes the plan's administrative costs.
- A comparison of the services each option provides.
- A comparison of the investment options available under each arrangement.
Alongside the analysis, the file needs the written disclosures: an acknowledgment that the firm and the professional are fiduciaries, a description of the services and the material conflicts of interest, and the written statement of the specific reasons the rollover is in this client's best interest. The DOL expects diligent and prudent effort to gather the plan data, and where the actual figures are not available it permits reasonable estimates based on the most relevant public data, such as the latest Form 5500 or benchmark data for a plan of that type and size.
The factor-by-factor split: what AI drafts, what a human signs
This is the heart of the workflow. Each required element has a data-and-drafting layer, which is machine work, and a judgment layer, which is human work. Confusing the two is how firms either drown in manual effort or, worse, let a tool make a determination it has no business making.
| Required element | What AI can gather and draft | What the licensed professional must do |
|---|---|---|
| Alternatives, including staying in the plan | Confirm whether the plan permits leaving assets in place; assemble the option set | Judge whether staying is genuinely viable for this client and weigh it honestly |
| Fee and expense comparison | Pull plan fees from disclosures or estimate from Form 5500 and benchmarks; build the side-by-side | Verify the estimates are reasonable and confirm the comparison is fair, not cherry-picked |
| Services comparison | Draft the description of services under plan versus IRA | Confirm the described services match what will actually be delivered |
| Investment options comparison | List and compare available investments under each option | Assess whether the difference actually serves this client's needs |
| Best-interest statement | Draft specific, client-referenced reasons from the analysis above | Own the determination and sign it; edit any reasoning that does not hold |
| Fiduciary and conflict disclosures | Assemble from Form ADV, Form CRS, and firm templates | Confirm accuracy and completeness for this relationship |
The line is consistent with every other part of an AI-assisted transition: the machine is trusted with gathering, comparison, and drafting, and the human is required for every determination and every signature. A best-interest statement that an algorithm generated and no advisor reviewed is not a best-interest analysis; it is a document, and an examiner will treat the difference as the whole point.
Where AI actually removes the pain
The reason this is worth automating is volume and repetition, not novelty. In a single breakaway, the same fee-comparison and disclosure-assembly work repeats across every client with an old employer plan, and each file pulls from the same handful of sources: the client's plan statements, the plan's public filings, the firm's ADV and CRS, and the advisor's own analysis. That is a document-extraction and assembly problem, which is exactly what FastTrackr's document intelligence is built to handle, reading plan statements and disclosures into structured data a template can populate.
Three specific gains show up. First, extraction: AI reads the plan disclosure or statement and pulls the fee and investment data instead of an analyst keying it by hand, and flags when the data is missing so the advisor knows to estimate rather than guess silently. Second, consistency: every rollover file in the book follows the same structure and pulls disclosures from the same current templates, which is what makes a later exam survivable, because uniform files are defensible files. Third, throughput: because an end-to-end advisor transition platform runs these concurrently rather than one at a time, the compliance documentation stops being the step that gates how fast the book can move. FastTrackr describes outcomes like moving large books in weeks rather than months, and those are FastTrackr's own reported results rather than independent findings, but the structural reason is real: repetitive, source-driven documentation is where automation compounds.
The same discipline extends past the rollover file into the rest of the transition's compliance surface, including the best-interest gaps that show up under Reg BI during a breakaway, which overlaps with PTE 2020-02 wherever a recommendation is involved. Firms that run transitions for a living treat this documentation layer as core rather than incidental, which is why transition consultants build it into their process from the start.
The retrospective review, and why clean files matter later
PTE 2020-02 does not end when the client signs. The firm has to run an annual retrospective review of its compliance with the exemption's conditions, document the results in a written report, and have a Senior Executive Officer certify it, with the review for a calendar year due by the middle of the following year. That review is only as good as the underlying files. If each rollover file is a bespoke, hand-built document with inconsistent structure and missing data, the retrospective review is an archaeology project. If every file was assembled to the same template with the same fields populated from the same sources, the review is a query.
This is the quiet, long-run argument for automating the documentation: not just that it is faster in the moment, but that it produces a records set the firm can actually review, certify, and defend later. A documented outcome from a real transition, like the one in FastTrackr's advisor transition case study, is worth more here than any feature claim, because the test of a rollover-documentation process is whether it holds up a year later under a senior officer's certification and, eventually, an examiner's questions.
What AI must never do in a rollover recommendation
The guardrail is the same and it is not optional. AI must not make the best-interest determination, because that is the fiduciary judgment the entire exemption exists to govern. It must not sign the disclosures or the best-interest statement, because the signature is the professional taking responsibility. It must not decide that a rollover is warranted when leaving assets in the plan is the better option, because that is precisely the conflicted call PTE 2020-02 is designed to check. And it must not present an estimate as a verified figure; when it estimates plan fees from public data, it should say so, and a human should confirm the estimate is reasonable. Inside those limits, AI does an enormous amount of useful work. Outside them, it manufactures exactly the compliance risk the firm was trying to avoid.
The bottom line for a compliance-minded firm
A breakaway moves accounts, but it also moves retirement money on the strength of the advisor's recommendation, and PTE 2020-02 is the rule that says a recommendation to leave a plan has to be documented, disclosed, and genuinely in the client's interest. The rule did not go away when the 2024 Fiduciary Rule did. The workload it creates is real, repetitive, and source-driven, which makes it a strong fit for AI on the drafting and assembly side and a strict no-go for AI on the determination and signature side. Get that split right and the rollover documentation stops being the bottleneck that slows a transition and starts being the clean, uniform, exam-ready record a firm is glad to have when the retrospective review comes due.
Frequently asked questions
Does PTE 2020-02 still apply after the 2024 Fiduciary Rule was struck down?
Yes. The 2024 Fiduciary Rule, which would have expanded the definition of who is a fiduciary, was effectively undone when the DOL withdrew its defense in late 2025 and federal courts in Texas finalized that outcome in early 2026. PTE 2020-02 is a separate, earlier exemption and remains in effect. Its core rollover obligations are unchanged: an advisor who provides ongoing advice for a fee and recommends rolling an employer plan into an IRA must perform and document a best-interest analysis and give the client a written explanation of the specific reasons the rollover is in their interest.
Which transition events actually trigger a PTE 2020-02 rollover analysis?
Only recommendations to move assets out of an employer retirement plan, such as a 401(k) or 403(b), into an IRA. A taxable brokerage account re-registering firm to firm through ACATS is a transfer, not a rollover. An IRA moving trustee to trustee is generally a transfer as well and does not by itself trigger the analysis. The event PTE 2020-02 governs is the advisor recommending that a client leave a plan. In a large book move, correctly sorting rollover recommendations from ordinary transfers is the first step, because a misclassified rollover is a file missing its required documentation.
What are the four factors an advisor must document for a rollover?
The advisor must evaluate and document the alternatives to a rollover, including leaving the money in the employer plan if that is permitted; a comparison of fees and expenses between the plan and the proposed IRA; a comparison of the services each option provides; and a comparison of the investment options available under each. The DOL expects diligent and prudent effort to gather plan data, and where actual figures are unavailable it permits reasonable estimates from the most relevant public data, such as the plan's Form 5500 or benchmark data for a comparable plan.
What parts of a rollover file can AI actually produce?
AI can read plan statements and disclosures to extract fee and investment data, estimate plan fees from public filings when the client's figures are missing, build the side-by-side comparisons of fees, services, and investments, draft a specific best-interest statement referencing the client's situation, and assemble the fiduciary and conflict disclosures from the firm's Form ADV, Form CRS, and templates. It can also flag missing data so nothing is silently guessed. What it cannot do is make the best-interest determination or sign the file, because those are the fiduciary acts the exemption exists to govern.
How does clean rollover documentation help with the annual retrospective review?
PTE 2020-02 requires the firm to conduct an annual retrospective review of its compliance, document it in a written report, and have a Senior Executive Officer certify it, with each calendar year's review due by the middle of the following year. That review depends entirely on the quality of the underlying files. Bespoke, inconsistent, hand-built rollover files turn the review into a manual reconstruction, while files assembled to a uniform template with the same fields from the same sources make the review a straightforward query, and make the eventual examiner conversation far easier to survive.
