An All Too Familiar Conformity Issue for Arizona Tax Professionals

“ADOR Outlines Executive Order and 2025 Tax Year Income Tax Forms,” Arizona Department of Revenue mailing list, January 22, 2026

As we prepare for the 2025 filing season, Arizona tax professionals who may be new to the Arizona specific quirks of annual conformity find themselves in a unique procedural posture for filing tax returns. However, as those who been in practice for many years in Arizona knows, a federal bill passing during the year with significant changes that, if adopted in full by Arizona would result in a substantial reduction in revenue, generally means that the conformity answer won’t likely be settled until after the April 15, 2027 unextended filing deadline.

Arizona Department of Revenue and Tax Forms

The Arizona Department of Revenue (ADOR) has issued individual income tax forms for the 2025 Tax Year that assume conformity with federal tax changes and incorporate state-level executive directives before they are officially enacted into law. This presents a technical non-conformity issue. Arizona’s tax law typically references federal values, such as adjusted gross income or itemized deduction computations, based on federal law as it stood on January 1, 2025. However, the federal “One Big Beautiful Bill Act” was not enacted until July 4, 2026, meaning the current ADOR forms anticipate federal law that post-dates Arizona’s current statutory reference date.

The Arizona Department of Revenue (ADOR) conventionally publishes tax forms based on this simplifying assumption, a practice that may seem unconventional to tax professionals outside of Arizona. This is because the state legislature historically does not adopt tax conformity—especially in years with a significant revenue impact—until the final budget bill is passed at the end of the regular legislative session. Based on past experience, this typically occurs well after the April 15 filing deadline. Most often either the Legislature and Governor agree to conform to all federal changes or only fail to conform to a limited number of items, so this default allows returns to be filed in a timely manner while limiting the number of taxpayers that must later amend their returns.

This article outlines the technical guidance provided by ADOR regarding the filing of 2025 returns and analyzes the specific provisions introduced by Governor Hobbs’ Executive Order 2025-15. Practitioners must understand the “file now, potentially amend later” stance taken by the Department and the specific tax relief measures currently embedded in the forms.

The “Presumed Conformity” Approach

Historically, ADOR has issued forms based on the assumption that the Arizona Legislature will conform to Internal Revenue Code (IRC) changes made in the prior year. For the 2025 tax year, this practice has been codified and expanded via Executive Order (EO) 2025-15.

The Department has explicitly stated that the 2025 forms reflect a “regular process of assuming conformity” to the IRC. The primary driver for this early release is the federal enactment of Public Law No. 119-21 (“H.R. 1”) in July 2025. Because Arizona law mandates Federal Adjusted Gross Income (FAGI) as the starting point for state income tax calculations, ADOR determined that failing to update forms to align with federal changes would disrupt the filing process for practitioners and taxpayers alike.

The specific concern cited by the Executive is the standard deduction. Since tax year 2019, Arizona has conformed to the federal standard deduction. The Governor’s office noted that H.R. 1 includes a higher standard deduction for tax year 2025. Without the Executive Order directing ADOR to update the forms to match this federal amount, approximately 90% of Arizona taxpayers (those claiming the standard deduction) would face immediate non-conformity, resulting in potential confusion and administrative rework.

The “Middle Class Tax Cuts Package” (The Governor’s Position)

Beyond standard IRC conformity, the 2025 forms include specific subtractions derived directly from Governor Hobbs’ Executive Order. The EO identifies these as the “Middle Class Tax Cuts Package”.

Practitioners should note that the 2025 forms currently allow for the following five specific provisions:

  1. Increased Standard Deduction: Matching the levels contemplated in H.R. 1,.
  2. Subtraction for Seniors: The EO specifies this as an “additional deduction of $6,000 for Arizonans aged 65 and older”. ADOR guidance confirms this subtraction is present on the released forms.
  3. Subtraction for Qualified Tip Income: Included as a specific line item subtraction,.
  4. Subtraction for Qualified Overtime Compensation: Included to provide relief for hourly workers,.
  5. Subtraction for Qualified Vehicle Loan Interest: This allows for the deduction of qualifying car loan interest,.

Incorporating these specific federal changes into the 2025 Arizona tax forms constitutes a departure from established practice. Typically, a bill solely advancing the Arizona conformity date to January 1, 2026, would not include these items in calculating Arizona taxable income.

Arizona utilizes its own distinct standard deduction, which is similar to, but not identical to, the federal standard deduction as it existed prior to the 2025 Tax Cuts and Jobs Act. Furthermore, the last four provisions listed do not impact federal Adjusted Gross Income (AGI) and are not treated as itemized deductions. Instead, they operate to reduce taxable income in a manner akin to the IRC §199A qualified business income deduction, which is never permitted for Arizona tax purposes. Their function is exclusively to be used in calculating federal taxable income pursuant to IRC §83.

It is important to recognize that while H.R. 1 reduced taxes for top earners and altered federal credits, ADOR’s forms only adopt the changes affecting FAGI and the specific relief measures directed by the Governor. Federal credits found in H.R. 1 that do not impact FAGI are not included on the Arizona forms.

Filing Guidance and Amendment Risks Given Subsequent Legislative Action

The most pressing question for CPAs is procedural: Should we hold returns until the Legislature convenes in January 2026?

ADOR’s guidance is unequivocal: No. Taxpayers should not wait to file.

However, this creates a distinct risk of bifurcation between the Executive’s forms and the Legislature’s eventual statutes. If the Legislature passes a conformity bill consistent with the forms released by ADOR, no further action will be required.

Conversely, if the Legislature does not approve these specific provisions, or passes a conformity bill that differs from the Governor’s EO (such as the provisions contemplated in SB 1106), taxpayers who utilized these subtractions may need to file amended returns.

The Legislature has passed a conformity bill that differs significantly from the Governor’s proposed Middle Class Tax Cut bill. The Governor subsequently vetoed the passed bill, and an override is highly improbable because the vote was strictly partisan. This situation arose following the issuance of the EO.

Penalty Relief Safe Harbor

Anticipating the potential need for mass amendments, ADOR has established a safe harbor for taxpayers caught between the Executive Order and Legislative action.

If a taxpayer files a 2025 return utilizing the provisions in the current forms, and those provisions are later deemed inconsistent with the final law passed by the Legislature:

  • ADOR will provide specific guidance on how to amend.
  • Taxpayers will not be subject to penalties or interest, provided the amended return is filed by October 15, 2027.

Conclusion

The release of the 2025 forms represents a strategic administrative move to prevent the “costly rework” of millions of returns regarding the standard deduction. However, by including the specific “Middle Class Tax Cuts” (senior, tip, overtime, and vehicle loan subtractions) prior to legislative approval, the Department has shifted the burden of monitoring legislative developments to the practitioner community.

We are advised to prepare 2025 returns using the forms as issued, taking advantage of the increased standard deduction and new subtractions where applicable. However, client communication letters should likely include a caveat regarding the pending legislative session and the remote possibility of a required amendment should the Legislature reject the Governor’s tax package.

Key Takeaway:

Per the Department of Revenue, tax professionals should:

  • File with Current Forms: Ensure all filings utilize the most up-to-date forms.
  • Monitor Legislation: Closely watch the 2026 Legislative Session for changes to Arizona’s conformity related to H.R. 1 and any subtraction modifications proposed by the Governor.
  • Utilize Safe Harbor: Rely on the October 15, 2027, safe harbor provision, which is anticipated to be necessary in the likely case where the final legislation differs from the Governor’s Middle Class Tax Cuts Proposal.

Prepared with assistance from NotebookLM.

The Statutory Void: How Sirius and Norwood Could Combine to Create Automatic SE Tax Liability for LLC Members

The Fifth Circuit’s recent decision in Sirius Solutions, L.L.L.P. v. Commissioner, No. 24-60240 (5th Cir. 2026), creates a strict textualist framework for the self-employment (SE) tax exclusion under I.R.C. § 1402(a)(13) for cases that would be appealed to the Fifth Circuit Court of Appeals.

While this decision provides a safe harbor for state-law limited partners, it simultaneously dismantles the functional analysis test that Limited Liability Company (LLC) members have historically relied upon to claim the exclusion. When the strict textual definition of “limited partner” adopted in Sirius is combined with the Tax Court’s precedent in Norwood v. Commissioner, a dangerous syllogism emerges: LLC members, lacking the specific state-law title required by the Fifth Circuit, may default to the general rule of inclusion regardless of their participation levels.

The General Rule of Inclusion: The Norwood Principle

To understand the risk posed by Sirius, one must first revisit the baseline established in Norwood v. Commissioner, T.C. Memo 2000-84. In that case, the Tax Court addressed the SE tax liability of a general partner in a medical supply partnership who spent only 41 hours on partnership matters during the tax year. The taxpayer argued that because his interest was passive, his distributive share should be exempt from SE tax.

The Tax Court rejected this argument, establishing a bright-line rule: unless a partner falls within the specific statutory exception for limited partners, their distributive share of trade or business income is subject to SE tax under Section 1402(a). The court held that the taxpayer’s “lack of participation in or control over the operations of [the partnership] does not turn his general partnership interest into a limited partnership interest”.

Crucially, Norwood established that for partners falling outside the specific Section 1402(a)(13) exception, economic reality regarding participation is irrelevant. The court stated: “That petitioner spent a minimal amount of time engaged in the operations of [the partnership] is irrelevant to this determination… [The distributive share is] subject to the taxes imposed by section 1401 on self-employment income… regardless of whether [the partner’s] involvement is passive or active”.

The Sirius Trap: Narrowing the Exception to State-Law Labels

Prior to Sirius, LLC members navigated the Norwood inclusion rule by arguing that, functionally, they occupied a position analogous to limited partners. However, the Fifth Circuit’s decision in Sirius Solutions ostensibly forecloses this “functional equivalent” argument by insisting on a strict adherence to the text of the 1977 statute.

In Sirius, the Fifth Circuit held that the “limited partner” exception applies only to a specific legal status. The Court declared: “When § 1402(a)(13) says ’limited partner,’ it is referring to a limited partner in a state-law limited partnership that has limited liability”. The opinion emphasizes that the “touchstone of a ’limited partner’ in 1977 was limited liability”, but inextricably links this liability protection to the specific entity format of a limited partnership.

While the Sirius court noted in a footnote that “we do not discuss whether members of another entity, such as an LLP or LLC, may also qualify for the limited partner exception”, the logic of the opinion presents a severe obstacle for LLC members. The Court expressly rejected the Tax Court’s approach of inquiring into the “functions and roles” of the partners. Instead, the Fifth Circuit vacated the lower court’s decision because it relied on an “erroneous passive investor rule”.

The Intersection: Why LLC Members May Be Left with No Defense

The combination of Norwood and Sirius leaves LLC members in a statutory void.

First, under Sirius, an LLC member cannot claim the Section 1402(a)(13) exclusion based on the plain text. An LLC member is not a “limited partner in a state-law limited partnership”. The Norwood court explicitly addressed the necessity of proper form, stating: “A limited partnership must be created in the form prescribed by State law”. Since an LLC is not a limited partnership under state law, the member fails the threshold definition established by the Fifth Circuit.

Second, the “functional analysis” escape hatch has been welded shut. In the past, an LLC member might have argued, “I act like a limited partner, so treat me as one.” However, Sirius explicitly condemns the IRS and Tax Court for applying a test that balances an “infinite number of factors in performing its ’functional analysis test’”. If the Fifth Circuit refuses to look at the “functions and roles” of a partner to grant the exclusion to a state-law limited partner, it is difficult to see how they would apply a functional test to grant the exclusion to an LLC member who lacks the requisite statutory title.

The Consequence: Irrelevance of Participation Level

If an LLC member cannot qualify as a “limited partner” under the Sirius definition, they revert to the general rule of Section 1402(a). Here, Norwood dictates the outcome.

In Norwood, the court held that because the taxpayer was not a limited partner, his income was subject to SE tax “regardless of whether [his] involvement is passive or active”. The court explicitly stated that the “passive activity rules under section 469 have no application in this case” regarding SE tax liability.

Norwood based its decision on the logic that the general rule of the first paragraph of IRC §1402(a) tells us partners pay self-employment tax on income from a trade or business carried on the partnership. It states:

(a) Net earnings from self-employment. The term “net earnings from self-employment” means the gross income derived by an individual from any trade or business carried on by such individual, less the deductions allowed by this subtitle which are attributable to such trade or business, plus his distributive share (whether or not distributed) of income or loss described in section 702(a)(8) from any trade or business carried on by a partnership of which he is a member (emphasis added);…

To avoid this result, a partner needs to look to the exceptions found enumerated in the remainder of IRC §1402(a), of which IRC §1402(a)(13) provides the limited partner exception found in this case.

While the Tax Court in Renkemeyer, Campbell & Weaver, LLP v. Commissioner, 136 TC 137, effectively reversed the Norwood analysis as applied to LLCs, the Fifth Circuit panel clearly states that the functional analysis the court turned to in Renkemeyer is not an appropriate inquiry to define the “limited partner, as such” in IRC §1402(a)(13).

Therefore, applying the Sirius logic to an LLC member results in the following position:

  1. The LLC member is not a “limited partner” because they are not in a state-law limited partnership.
  2. The LLC member cannot use a “functional analysis” to prove they are a limited partner, as that test is erroneous.
  3. Without the Section 1402(a)(13) exclusion, the member is subject to the general rule of Section 1402(a).
  4. Under the general rule (Norwood and the plain text of IRC Section 1402(a)’s first paragraph), the member’s distributive share is fully subject to SE tax, even if they spent zero hours working for the entity.

Conclusion

While Sirius Solutions is a victory for traditional limited partnerships, it creates a high-stakes environment for LLCs electing partnership taxation. By defining “limited partner” strictly by reference to state-law limited partnerships and rejecting functional inquiries, the Fifth Circuit may have inadvertently subjected all LLC members to the Norwood regime: automatic SE tax liability on all trade or business income, regardless of whether the member is a passive investor or an active manager. As the Sirius court noted, “state law creates legal interests… but the federal statute determines when and how they shall be taxed”. Without the specific state-law interest of a “limited partnership,” LLC members currently lack a judicially recognized shield against self-employment tax in the Fifth Circuit.

Prepared with assistance from NotebookLM.

Arizona Proposed Tax Conformity and Reform: A Technical Analysis of HB 2153 and SB 1106

The Arizona Legislature is currently moving two identical bills, HB 2153 and SB 1106, which propose substantial modifications to the Arizona Revised Statutes (A.R.S.) regarding income taxation in reaction to the enactment of the One Big Beautiful Bill Act. But despite rapid movement this week, it make take quite a while before we get a complete answer to how Arizona will deal with conformity this year.

While introduced separately, the legislature intends to merge these provisions into a single enactment. The legislation updates Arizona’s conformity date with the Internal Revenue Code (IRC), overhauls the standard deduction calculation, and introduces significant new subtractions from gross income for retirement and labor-related income. However the proposal would add a new nonconformity provision related to itemized deductions for state and local taxes.

While the bill is likely to receive approval from both the House Ways and Means Committee and the Senate Finance Committee in a rare joint committee meeting on January 14, the Arizona Capitol Times reported on January 13 that the bill faces a likely veto when it arrives at the Governor’s desk, while the publication also reported Senate President Warren Petersen stated the bill should arrive on her desk in days.

The Governor does not want to deal with conformity to provisions other than the ones she discussed in her Executive Order released in November until an agreement is reached on the budget for this session. So both the Executive Order and this bill appear to be opening proposals in what seems likely to become an extended battle over budget issues this session.

Even if the bill is vetoed, it is likely that eventually the parties will agree to some form of compromise that includes many of these details, so it is helpful to understand what is in this bill. Just don’t be surprised if we don’t get a bill passed and signed into law until well after the April 15 deadline.

Technical Analysis of the Bill

The following technical analysis details the additions, deletions, and modifications contained in the legislation.

I. Internal Revenue Code Conformity (A.R.S. §§ 42-1001, 43-105)

The legislation updates the definition of the “Internal Revenue Code” for Arizona income tax purposes.

  • Updated Reference Date: For taxable years beginning from and after December 31, 2025, the definition of the IRC is updated to mean the Code as in effect on January 1, 2026.
  • Retroactive Provisions: The bills incorporate provisions that became effective during 2025 with the specific adoption of their retroactive effective dates.
  • Historical Conformity: The bills strike out the specific conformity definition for the 2014–2015 period (Subsection K), effectively rolling the historical window forward.

II. Standard and Itemized Deductions (A.R.S. §§ 43-1041, 43-1042)

Perhaps the most structural change involves the decoupling of the Arizona standard deduction from fixed state-statutory dollar amounts.

A. Standard Deduction Coupling (A.R.S. § 43-1041(A))

Practitioners should note that the specific statutory amounts (e.g., $12,200 for single filers) and the accompanying inflation adjustment mechanisms have been deleted.

Instead, the Arizona standard deduction is now statutorily defined as “the amount of the federal basic standard deduction determined pursuant to section 63 of the internal revenue code for the taxpayer’s filing status”. This creates an automatic lockstep with federal inflation adjustments, removing the need for separate state-level inflation calculations previously mandated under the now-repealed Subsection H.

B. Charitable Contribution Adjustment (A.R.S. § 43-1041(H))

For taxpayers electing the standard deduction, the “charitable add-on” calculation (the additional amount allowed beyond the standard deduction) changes significantly for taxable years beginning from and after December 31, 2025:

  • Old Law: 25% of total charitable deductions allowed, with the percentage annually adjusted by a factor determined by the rate of inflation.
  • New Law: The add-back is equal to 100% of the taxpayer’s charitable contributions as defined in IRC § 170©, but it is subject to a strict statutory cap:
    • $1,000 for single or married filing separately.
    • $2,000 for married couples filing jointly.

C. Itemized Deduction for SALT (A.R.S. § 43-1042(D))

A new provision is added regarding the deduction for state and local taxes (SALT). In lieu of the federal itemized deduction amount allowed under IRC § 164(b)(7), an Arizona taxpayer may deduct up to $10,000 of that amount for state and local taxes. If the bill is enacted into law, Arizona is effectively keeping the “old law” (TCJA) version of the state and local tax deduction, deciding not to conform to the temporary higher deduction allowed on the federal return.

III. Subtractions from Gross Income (A.R.S. §§ 43-1022, 43-1030)

The legislation introduces several new subtractions from Arizona gross income, generally effective for taxable years beginning from and after December 31, 2024. These provisions implement two of the four deductions added to IRC §63 available regardless of whether or not a taxpayer itemizes deductions, but which do not reduce a taxpayer’s federal adjusted gross income. The law provides a substitute for one other of the deductions and quietly declines to adopt the fourth.

A. Labor and Compensation

The two federal Section 63 deductions proposed for adoption into Arizona tax law pertain to compensation.

  • Qualified Tips: Taxpayers may subtract the amount of qualified tips received that are deducted under IRC § 224. (ARS §43-1022(31))
  • Overtime Compensation: Taxpayers may subtract qualified overtime compensation deducted under IRC § 225. (ARS §43-1022(32))

B. Retirement Income (A.R.S. §§ 43-1022(36) and 43-1030)

The bills introduce a complex subtraction matrix for retirement income involving Roth IRAs and pension distributions.

  1. Roth IRA Contributions: A subtraction is allowed for contributions to a Roth IRA (IRC § 408A) made during the taxable year, provided they were not deducted from federal AGI.
  2. Pension/Retirement Distributions (New § 43-1030): Taxpayers aged 60 or older may subtract distributions from qualified pension or retirement accounts.
  3. Aggregate Limits: The combined subtraction for Roth contributions and pension distributions is capped at:
    • $6,000 for single/head of household/MFS.
    • $12,000 for married filing joint.
  4. Means Testing (Phase-out): The subtraction under § 43-1030 (distributions) is reduced by 6% of the amount the taxpayer’s Arizona gross income exceeds:
    • $75,000 (Single/HOH/MFS).
    • $150,000 (MFJ).

This appears to be in lieu of Arizona adopting the over 65 subtraction that the Governor had proposed in her Executive Order.

C. No Tax on Car Interest Deduction Not Adopted

The bill contains no text related to the no tax on car interest deduction added as part of the One Big Beautiful Bill, which serves to decouple Arizona’s taxable income computation from this federal provision.

Arizona tax law starts with federal adjusted gross income and allows itemized deductions by reference to the federal itemized deductions, which means simply updating the IRC conformity date would not bring in a provision that allows a deduction solely in the computation of taxable income as part of IRC §63, such as the deduction for qualified business income under IRC §199A.

So by its silence on the issue, the bill provides that Arizona law will not conform to the federal deduction for certain interest paid to acquire a qualified vehicle.

D. Family and Education

  • Dependent Care: Taxpayers may subtract child and dependent care expenses (IRC § 21) that exceed the federal credit amount received. (ARS §43-1022(34))

  • Adoption Costs (Post-2025): For taxable years beginning after December 31, 2025, the subtraction limit for adoption costs increases from $3,000 to:

    • $5,000 (Single/HOH).
    • $10,000 (MFJ).

    This provision is already part of Arizona tax law, enacted in 2025’s session to take into account the expiration of a higher temporary increase in the allowable adoption expense.

  • 529A/530A Distributions: A subtraction is added for distributions from accounts established pursuant to IRC § 530A (Trump Accounts). (ARS §43-1022(33))

IV. Credits and Other Provisions

A. Dependent Tax Credit (A.R.S. § 43-1073.01)

The bills amend the dependent tax credit amounts. For dependents under seventeen years of age, the credit increases from $100 to $125. The credit for dependents 17 and older remains $25. The phase-out thresholds remain unchanged.

B. Scholarship Granting Organizations (New Chapter 18)

The legislation adds Chapter 18, titled “Scholarship Granting Organizations”.

  • Federal Election: Arizona elects to participate in the federal tax credit established by IRC § 25F.
  • Certification: The Department of Revenue must certify SGOs that meet IRC § 25F requirements.
  • Effective Date: Participation and scholarship issuance under this chapter begin for taxable years from and after December 31, 2026.

V. Summary of Effective Dates

  • Retroactive to Tax Years Beginning After Dec 31, 2024:
    • New subtractions for tips, overtime, dependent care, and retirement (Roth/Pension).
    • IRC Conformity update (Jan 1, 2026 reference).
    • Standard Deduction coupling to IRC § 63.
  • Effective for Tax Years Beginning After Dec 31, 2025:
    • Charitable contribution standard deduction adjustment.
    • Revised adoption cost subtraction.
    • IRC § 530A (Trump Accounts) (This is the practical effective date, as federal law does not allow any funds to be deposited into a Trump account until July 4, 2026, thus no distributions are possible before that date. Technically this provision takes effect for tax years beginning after December 31, 2024 but it simply could not apply to 2025 returns.)
  • Effective for Tax Years Beginning After Dec 31, 2026:
    • Scholarship Granting Organization provisions. (Another practical effective date exists, as the related federal tax credit will not take effect until 2027. Technically, all of these provisions are effective immediately, with the exception of ARS §43-1803, which pertains to organizations issuing SGO scholarships.)

Prepared with assistance from NotebookLM.

The Arizona Middle Class Tax Cuts Package: Operational Considerations for the 2025 Filing Season

Governor Katie Hobbs issued Executive Order 2025-15 on November 20, 2025, titled “Cutting Taxes for Hard-Working Arizonans”. This action initiates a series of prospective changes to the Arizona state income tax structure, primarily aimed at middle-class taxpayers. Given that the Executive Order (EO) directs the Arizona Department of Revenue (ADOR) to update state tax forms for the 2025 tax year, tax professionals must understand the technical directives of the order and the subsequent legislative action required to ensure the validity of these planned benefits.

Executive Order 2025-15: Directives and Scope

The core function of Executive Order 2025-15 is to mandate that ADOR update the State of Arizona tax forms to enable taxpayers to utilize the proposed Middle Class Tax Cuts Package when filing their 2025 tax returns.

The EO directs ADOR to implement two distinct categories of adjustments:

  1. Standard Deduction Update: The tax forms will be updated to include the higher standard deduction passed into law by H.R. 1. This brings Arizona tax forms into conformity with portions of H.R. 1 that benefit middle-class Arizonans. Specifically, the state standard deduction is set to increase from the previous $15,000 amount to $15,750 for single filers and $31,500 for joint filers.
  2. Prospective Instructions for Legislative Subtractions: The ADOR is also directed to include prospective instructions for Arizonans to take advantage of several new deductions. These instructions are included in anticipation of the legislature subsequently codifying these proposed measures into law.

It is important to note that any further tax conformity provisions, particularly those affecting corporations and high earners, are intended to be negotiated separately in the FY27 budget, consistent with the standard procedure for legislation that carries fiscal impacts.

Specific Proposed Subtractions and Deductions

The Middle Class Tax Cuts Package introduces several specific income subtractions designed to reduce Arizona taxable income:

  • Senior Deduction: An additional deduction of $6,000 for Arizonans over the age of 65.
  • Tipped Income Subtraction: Arizonans will be allowed to deduct tipped income from their state taxes. For instance, a hypothetical taxpayer’s return showed a $5,000 subtraction for tip income.
  • Overtime Income Subtraction: Arizonans will be allowed to deduct eligible overtime income from their state taxes. A hypothetical example illustrated a $3,000 subtraction for overtime wages.
  • Car Loan Interest Subtraction: Arizonans will be allowed to deduct interest for car loans on new, American-made vehicles from their taxes. A hypothetical taxpayer was shown benefiting from a $4,000 subtraction for car interest.

These combined subtractions and the increased standard deduction significantly impact the calculation of Arizona Taxable Income. In a modeled scenario, a 68-year-old single filer with a Federal Adjusted Gross Income (AGI) of $75,000 saw their Arizona Taxable Income drop from $60,000 (using the $15,000 standard deduction) to $41,250 (using the $15,750 standard deduction and $18,000 in additional subtractions).

Implications for Arizona 2025 Income Tax Returns

The Executive Order places ADOR in the position of issuing tax forms that incorporate subtractions that have not yet been statutorily enacted.

For tax professionals preparing 2025 returns, the primary operational challenge lies in the distinction between the standard deduction increase and the new subtractions:

  1. Standard Deduction: The increased standard deduction ($15,750 single / $31,500 joint) is included in the updated forms because it was “passed into law by H.R. 1”.
  2. Middle Class Subtractions: The deductions for tipped income, overtime, senior status, and car interest are based on “prospective instructions”. This means that while ADOR has been directed to enable these subtractions on the forms, their legal efficacy depends entirely upon the subsequent legislative action.

If the legislature fails to codify the Middle Class Tax Cuts Package, taxpayers who rely on the deductions specified in the ADOR’s prospective instructions could face deficiency assessments, as the subtractions would lack statutory authority.

The Urgency of Legislative Codification

Governor Hobbs has called upon the state legislature to immediately pass the Middle Class Tax Cuts Package at the beginning of the session. This urgency stems from the need to provide tax certainty to everyday Arizonans.

Crucially, the vast majority of taxpayers will file their 2025 returns prior to the anticipated completion of the FY27 budget. If the legislature delays the codification of the tips, overtime, senior, and car interest subtractions until the budget process concludes, those taxpayers who file early and claim these deductions, based on ADOR’s instructions, risk filing an erroneous return if the provisions are altered or rejected.

Therefore, for CPAs and EAs seeking to ensure clients maximize benefits without incurring future amendment requirements, the prompt codification of the Middle Class Tax Cuts Package into state law by the legislature is essential. Codification at the start of the session would retroactively grant the necessary statutory authority to the subtractions, aligning the law with the tax forms ADOR has been directed to issue under EO 2025-15.

The filing certainty for Arizona taxpayers regarding these proposed subtractions is currently akin to building a bridge that is authorized but not yet anchored; while the path forward is visible on the forms, its structural integrity relies entirely on the timely arrival of statutory legislative support.

Prepared with assistance from NotebookLM.

Sometimes Memorization Just Ain’t Enough (With Apologies to Patty Smyth for Borrowing Her Title)

In a recent article published by Futurism, Jad Tarifi, the founder of Google’s inaugural generative AI team, posited a cautionary stance regarding careers in law and medicine, asserting that these professions primarily entail the memorization of information. This assertion prompted a critical examination of prevalent misconceptions concerning the true nature of knowledge and skill acquisition within these domains.

A compelling illustration of this misunderstanding is evidenced by the consistent erroneous responses from three distinct AI large language models (LLMs) when queried about specific provisions within the One Big, Beautiful Bill Act (OBBBA) designed to reduce an individual’s adjusted gross income. As I documented in a prior blog entry, I, a certified public accountant (CPA) without additional legal training, and certainly not possessing comprehensive recall of the entire OBBBA, was immediately able to identify the flaw in the models’ conclusions. This occurred despite the theoretical premise that these models “knew” the entirety of the Internal Revenue Code (IRC), along with all pertinent binding guidance, and had unfettered access to the complete legislative text that modified the aforementioned law.

Within tax law, rote memorization has never been, nor will it ever be, the sole method for ascertaining legal principles and their application. Instead, a meticulous reading of the law is imperative, necessitating careful attention to its enumerated items, cross-references, and the subsequent tracing of those references (which may lead to further cross-references) to construct a comprehensive understanding of the statute. Furthermore, the process involves distinguishing unambiguous statutory provisions, which dictate a singular interpretation, from those exhibiting ambiguity, which require interpretation. This interpretive process relies upon established canons of statutory construction developed over time.

Subsequent to this initial statutory analysis, the practitioner must then explore other existing interpretive binding official guidance, assess whether the law has undergone changes since such guidance was issued, and finally, consult expert discussions from third-party sources and non-binding official guidance. This comprehensive approach yields potentially supportable interpretations of the matter at hand, alongside an assessment of the likelihood that any given interpretation would ultimately be accepted by an IRS agent, an appellate conferee, or the highest court to which the case might escalate.

The fundamental flaw in the analyses provided by ChatGPT, Gemini, and BlueJ in this instance was their failure to commence with an isolated examination of the law. Instead, they prematurely resorted to other sources, neglecting to analyze the memorized legal text and the legislative amendments in isolation. Moreover, it remains unclear whether such isolated analysis is truly feasible for these models given their operational paradigms.

As previously detailed, the OBBBA provisions concerning the exemption of tips, overtime pay, and car loan interest from taxation were appended as deductions to Section 63 of the IRC. This classification renders them deductions taken along with either the standard deduction or itemized deductions in computing taxable income, rather than deductions that reduce adjusted gross income. As elaborated upon in the earlier article, deductions utilized in the computation of adjusted gross income would be allocated to Section 62.

While the models had purportedly memorized the entire structure of the IRC, theoretically imparting this knowledge, they failed to undertake the crucial step that tax CPAs and attorneys should perform at this juncture: consulting the statutory text and following cross-references to gain an initial understanding of the IRC’s provisions as amended pertaining to this issue. Instead, they were diverted by other related information contained within their models and acquired from the web during data updates to fill in for developments taking place after the end of time covered by their training. This data included articles quoting Congressional sources that correctly noted these items were deductible regardless of whether a taxpayer itemized deductions. Furthermore, the training data encompassed years of articles that referred to various prior specific items deductible irrespective of itemization as “above the line” deductions. Their data also indicated that “above the line” signified a deduction used in computing adjusted gross income. None of this memorized data was factually incorrect, though much of it proved irrelevant in this specific context.

Consequently, the models quickly identified a high correlation between the terms “deductible even if the taxpayer does not itemize deductions” and a deduction being categorized as “above the line,” as well as a nearly one-to-one correlation between a deduction being referred to as “above the line” and its use in computing adjusted gross income. Again, all of this derived from the memorized information was accurate.

However, at this juncture, the analysis fell apart due to an incorrect inference. The models utilized all data indicating these deductions were available to both itemizing and non-itemizing taxpayers (which was entirely correct) to infer that these deductions were “above the line” and utilized in computing adjusted gross income (absolutely not correct). This inference was reinforced by the historical commentary present within their vast memorized data. Nevertheless, it was an inference that proved completely erroneous.

The question then arises: why did the models’ access to more current sources not challenge their conclusions? This highlights the fallibility of many human authors who were expeditiously producing commentary on the bill. Numerous such articles, some originating from highly reputable sources, incorrectly labeled these provisions as “above the line” deductions.

While the precise reasons for these authors’ misclassification remain speculative, it appears evident that they did not base their initial articles on a direct analysis of the bill itself and the Internal Revenue Code for this issue. It is plausible that many made the same inference that the LLMs derived from their extensive memorized data—namely, that historically, Congress had enacted provisions for items deductible even without itemization as deductions used to compute adjusted gross income. Others, encountering previously published articles making the claim that these deductions reduced adjusted gross income, simply relied upon those sources. Furthermore, I suspect that some authors, operating under time constraints, posed the question to LLMs, which, based on their initial model assumptions, provided the erroneous result.

In any event, the articles discovered by the LLMs analyzing the law served to corroborate their models’ initial assumptions, which were rooted in their extensive memorized data, thereby creating a significant feedback loop.

My approach to the new legislation mirrored that of previous laws, with the notable addition of Google’s NotebookLM, an AI application primarily reliant on user-provided sources. The notebook I established contained exclusively the tax provisions of the statute. As I systematically reviewed each section of the bill, I utilized NotebookLM to generate summaries of the respective sections, while simultaneously conducting my own independent review to ensure comprehensive understanding and prevent oversight during rapid analysis of the law.

NotebookLM’s analysis was largely confined to the provided document, offering little beyond the assertion that each provision modified Section 63 since that was all that was available to the application. Consequently, I cross-referenced Section 63 to ascertain the specific implications of such modifications. As I knew Congress had, on a rare occasion, introduced an additional “below the line” deduction by placing it in this Section, accessible regardless of whether a taxpayer itemized deductions. They had decided to do this multiple times in the new bill.

A precedent for such a deduction can be found in the Qualified Business Income (QBI) Deduction under IRC §199A, enacted by the Tax Cuts and Jobs Act of 2017. This provision has remained in effect and its application was extended as part of the OBBBA.

Posing inquiries to the models regarding the placement of these deductions within Section 63, akin to the QBI deduction in 2017, frequently resolved the models’ prior limitations. This shift in focus directed their attention to articles detailing that specific deduction and highlighting the impact of its placement in Section 63 rather than Section 62. The models subsequently confirmed that the bill text indeed added these deductions to Section 63 and, likely drawing from the newly focused articles, concluded that these deductions do not reduce adjusted gross income.

Even if their approach remained flawed, they did finally arrive at a correct solution.

I am not an attorney and do not claim to be one. However, I would be astonished if similar problems didn’t frequently emerge when parties depend on LLMs for tasks typically handled by law students who would become attorneys, whom Mr. Tarifi suggests should discontinue their legal studies.

Mr. Tarifi undoubtedly possesses far greater knowledge of LLMs and AI than I ever will. Yet, I question whether proponents of AI solutions truly grasp the issues they aim to resolve, or if they merely presume to understand them. The immediate availability of the entire IRC and all supporting documentation within an LLM’s model is genuinely incredibly useful and will revolutionize our approach to tax research and analysis. Nevertheless, I hope I have demonstrated how mere rote memorization is insufficient to perform the job being done today—nor do I observe signs that newer models are progressing beyond memorization to conduct a comprehensive analysis of new laws and developments.

Large Language Models (LLMs) represent a significant advancement that professionals in tax practice must comprehend to maintain relevance. However, similar to the introduction of computers in the 1960s and 1970s and subsequent machine learning enhancements beyond initial automation of routine preparation tasks, LLMs are not anticipated to be the technology that renders tax professionals obsolete.

An Interesting Error from LLMs in Tax Research That Does Not Seem to Be a Hallucination

I was experimenting with three LLMs on a tax research issue recently. I asked them to identify deductions newly available to individuals without a business or rental activity that are deductible in computing adjusted gross income (AGI).

ChatGPT, Gemini, and BlueJ (a paid service for tax professionals) all identified the following:

  • No tax on tips
  • No tax on overtime
  • No tax on interest on car loans

As I’ll show, this answer is wrong. But first, consider this: they all arrived at the same incorrect answer. While we know that LLMs can hallucinate, this isn’t likely a hallucination (an answer created from thin air). If all three models were hallucinating, they would be highly unlikely to invent the exact same error.

The key to the error lies in the distinction between IRC §62 and §63. If you’ve taken my course on the OBBBA, you’ll recall I noted that these specific deductions are routed through §63 (computation of taxable income), not §62 (computation of adjusted gross income). The most well-known §63 deduction is the QBI deduction under §199A-it reduces taxable income on Form 1040 but does not reduce AGI. Because it doesn’t reduce federal AGI, it also doesn’t reduce Arizona taxable income, a crucial point for us here in Arizona.

So, how did all three AIs get the same wrong answer? It’s simple: a large number of human authors, whose material was used in the models’ training data or accessed via web searches, made the same mistake. Because these new deductions were touted as being available even to non-itemizers, many authors assumed they were “above-the-line” deductions and described them as such. However, a review of the statute makes it clear: the OBBBA added these provisions to §63 but made no change to §62.

It’s true that the LLMs didn’t consult the text of the law; they don’t perform legal analysis. Rather, they synthesize the analysis that others have prepared and published, giving extra weight to “high-quality” sources. In this case, a large percentage of human analysts made the same mistake.

There are understandable reasons for this. Bill proponents repeatedly noted the deductions were available even if a taxpayer did not itemize. Many general-purpose financial publications quickly released articles stating these deductions were “above the line.” They presumably equated being able to take a deduction without itemizing with it being an “above-the-line” deduction-an association that was generally, but not always, true before the TCJA and OBBBA.

From personal experience, I know producing an analysis of a new tax bill is done under extreme pressure. Reading federal legislation is messy; the bill only shows the amendments, not the law in its final, consolidated form. It’s easy to miss the significance of a deduction being added to §63 instead of §62.

This distinction is critical because AGI impacts numerous calculations. For states like Arizona that start with federal AGI, these deductions won’t reduce state taxable income. Furthermore, AGI affects tax thresholds (like phase-outs and deduction floors) and even non-tax items like the IRMAA calculation for Medicare premiums.

The takeaway is that relying on a human-written analysis is no guarantee of correctness, either. I have sat through continuing education webinars on the OBBBA that contained this exact error. At this point, the error is so common that it has likely “infected” other human authors, who repeat it after hearing it from multiple sources.

I fully understand how this error was made and could easily have made the same mistake myself (and have likely made other mistakes).  But unless you are working directly from source materials you have to always remember that there could be flaws in the analysis you are reading-and sometimes these errors become self-reinforcing when authors have seen or read previous analyses.

Gemini “Thought” Process for Deep Research – How It Produced the Correct Analysis While Continue to “Think” It Was Dealing with Fabricated Detail for a Non-Existent Law

I have been looking at various AI sources for dealing with tax updates, including new law such as the One Big Beautiful Bill that was enacted on July 4, 2025 (I did mistakenly refer to the signed into law date in my prompt as July 1, 2025 but that did not cause the issue, though you will notice that date in Gemini’s response to my query). And I noticed one thing about Gemini when I asked about the Trump Accounts in it, something I knew had changed dramatically in various versions of the bill. When I told Pro 2.5 to give me information about the bill signed into law on July 4, 2025 it replied that no such bill exists. That didn’t surprise me, as I didn’t expect it training to be that up to date.

So I next turned to Deep Research. The report it generated was the best one I’ve gotten from any AI system (including BlueJ, a commercial tax professional product) but in looking at its “thoughts” I found it initially concluded that the bill did not exist, then did its web search, found a lot of high quality sources that did talk about it, but it continued to refer to this information as fabricated in the thought process (the bill didn’t exist). Nevertheless, it produced summary that the thought history shows it believes was fabricated, but which never is labeled as such in the output.

So I decided to pose the question to Gemini 2.5 Pro about why that had happened, and it’s explanation, reproduced below, is an interesting look at how that system works. And why the “thoughts” are more of a look at the internals and not necessarily reflecting a “belief” in any way you might interpret that word.

Gemini’s Response Follows:

That is an excellent and highly insightful question. Your observation gets to the very heart of how advanced AI systems like Gemini, particularly with features like Deep Research, function. You are correct in your core assumption, but there’s a fascinating nuance to why you see that “cognitive dissonance” in the process.

Here’s a breakdown of what’s happening, addressing your specific question about why the model seems to maintain its initial belief in the “Thoughts” while producing a correct final output.

The Two “Minds” of the Operation: Base Model vs. Research Agent

Think of the process not as a single mind that changes its opinion, but as a multi-step workflow involving different specialized components or “agents.”

  1. The Planner (Based on Gemini 1.5 Pro’s Static Knowledge):
    When you initiate Deep Research, the first step is for the base model (Gemini 1.5 Pro) to understand your request and formulate a research plan.
    • Initial Check: The model’s first instinct is to check its own internal, static knowledge base. In your scenario, its training data cuts off before July 2025.
    • The First “Thought”: When it checks for “HR1 signed into law on July 1, 2025,” its internal knowledge returns a definitive “no match found.” This is a factual statement based on its training data. This is where the thought you observed—“the bill does not exist”—originates.
    • The Plan: Crucially, the model doesn’t stop there. It recognizes that its knowledge is limited by time. The plan it generates will involve steps to overcome this, such as:
      • Search for information about a bill designated HR1 in 2025.
      • Verify if HR1 was signed into law in July 2025.
      • Find authoritative sources like congress.gov, news reports, and legal analyses for this bill.
      • Synthesize the findings into a comprehensive analysis.
  2. The Researcher (The Tool-Using Agent):
    This is the part of the process that executes the plan. It autonomously runs Google searches, reads the results from the credible sources it finds, and gathers the raw data.
    • Finding New Evidence: In your hypothetical scenario, this agent finds a wealth of information about your fictional HR1 from authoritative sources published after its knowledge cutoff. It discovers the bill’s text, news reports on its signing, and analyses from think tanks.
    • Data Collection: This new, contradictory information is collected. The system notes the sources and the content. This is the second part of the “Thoughts” you see, where it lists the sources it found and the information it extracted.
  3. The Synthesizer (The Final Report Generator):
    This is the final and most important step. The system now has two conflicting sets of information:
    • Internal Knowledge (from the Planner): “This bill doesn’t exist.”
    • External Knowledge (from the Researcher): “Multiple, fresh, authoritative sources confirm this bill exists and here are its complete details.”

This is the key to your question: The system is not designed to “believe” one thing over the other in a human sense. It is designed to **fulfill the user’s request using the best available information.**The ultimate goal is to provide an analysis of the bill. The information retrieved from the live searches is infinitely more relevant, current, and useful for that task than its own outdated internal knowledge. Therefore, the synthesis process is architected to prioritize and build the final report based on the fresh data gathered by the Researcher. The initial (and now invalidated) thought from the Planner is correctly discarded as an artifact of the research process, not a fact to be included in the final product.

Why Does the “Thought” Linger?

The “Thoughts” section is a transparent log of the agent’s work process. It’s showing you its homework.

  • It shows the initial state (Premise: Check for HR1. Result: Not found in my knowledge base.).
  • It shows the actions taken (Action: Search Google. Results: Found 10 credible sources…).
  • The log isn’t a “stream of consciousness” that constantly edits itself. It’s a record. So, the initial premise remains logged even after it has been superseded by new evidence.

You are essentially seeing the “ghost in the machine”—the base model’s initial, correct-but-outdated assessment—before the research agent provides new, more relevant data for the final output.

Analogy: The Diligent Librarian

Imagine you ask a librarian (with a perfect memory up to 2024) for a book published yesterday.

  1. Initial Thought: The librarian thinks, “I have memorized every book in this library’s catalog. That book is not in my memory. Therefore, it does not exist in my catalog.”
  2. Action: However, they are a diligent librarian. They don’t stop there. They say, “Let me check the ‘New Arrivals’ cart from today’s delivery.”
  3. Discovery: They walk over to the cart, find the book, and read its back cover and table of contents.
  4. Final Answer: They walk back to you and give you a perfect summary of the new book. They don’t start by saying, “Well, I first thought this book didn’t exist…” They give you the answer you asked for, based on the new information they actively sought.

The “Thoughts” in Deep Research let you watch the librarian walk to the cart. The final output is the summary they give you at the end. Your observation is so keen because you noticed the librarian’s initial, momentary flicker of doubt before they completed their process.

Senate Finance Committee Initial Draft of Reconciliation Bill Tax Provisions

I’ve uploaded a (primarily) NotebookLM section by section summary of tax related provisions found in the Senate Finance Committee draft issued on June 16, 2025, using the law text and the committee’s section by section summaries as inputs. While I’ve roughly reviewed each section’s explanation, I have not done a deep dive on confirming treatments since this clearly won’t be the final bill.

For those note aware, NotebookLM is a Google notes/study web application that uses specific sources uploaded to each notebook as the inputs to answer questions, generate summaries, FAQs and study materials on the provided items. The inputs given were:

Finance Committee Proposed Legislative Text

Section by Section Details in Summary Form

My document was generated by making a standard inquiry of NotebookLM on a bill section by section basis. That helps focus the program on the items in that section and reduces (but does not eliminate) the risk that it will find and misinterpret irrelevant items in the texts about each section.

Here is the query I used, changing the section number for each run.

Please write a summary of Section 71304 of the proposed legislation.  Outline the provisions found in current law, how the bill would change the law going forward, and any Internal Revenue Code provisions impacted by the law change.  You are a CPA experienced in tax matters writing an article to explain this provision to other experienced tax professionals.  Try to maintain a neutral position on any policy related matters, concentrating primarily on the mechanics of the operation of the old and new law.  Do not discuss IRC Section 15 revisions in this article. Just include the article, with no text before or after the article you generate. The title of the article should only be the word “Section” followed by the section number and then the formal title of the section.

A couple of notes on the last few sentences. The Section 15 discussion in both the bill and the summary applies generally to the bill, so NotebookLM wanted to keep bringing it up every time even though it really only impacts a few provision. I kept it in the first section discussion on rates, but then suppressed it. I also didn’t really want the “chatty” description of what I had asked (so thus the suppress text before and after the article) and the last couple of sentences just helped to get more consistent headers and titles.

Here is the entire document that was created on this bill. Again, use this as a rough guide and it some issue is crucial for your client, please go directly to the sources noted above. Any time something is “summarized” by definition some things get left out, and sometimes those details are important in your client’s situation. As well, errors do creep in for various reasons, so confirming any editorial source (including CCH and RIA’s) is an important step before relying on any analysis.

Senate Finance Commmittee Tax Bill NotebookLM DocumentDownload

The One, Big, Beautiful Bill Title XI Tax Provisions as Passed by the Ways & Means Committee

I decided to post the following for various reasons at this point. First, since the One, Big, Beautiful Bill has already had some amendments made to Title XI by the House Rules Committee (though the full updated bill text is not available as of the morning of May 24) there’s little reason to refine this document. Second, I have found it useful as a starting point for answering questions about the bill (being able to find relevant items in the bill text and JCT report). And, third, it does demonstrate the usefulness of NotebookLM for getting on top of these sorts of things, as well as gave me a good learning experience in how to deal with new law text (the JCT report was really useful in being able to handle the bill, since the latter only comes with the changes to the IRC, not the full revised IRC provision in context which is often needed to understand what Congress is doing).

So I am posting the analysis I compiled of the bill using NotebookLM, the bill text that was approved by the Ways & Means Committee, and the JCT report on the various provisions in that bill. I added a section on the proposed changes to PL 86-272 as well though that was added by a different committee and is not part of Title XI.

The way I handled it was to craft a prompt to ask for an analysis dealing with specific areas of each section in the bill, working from the section list in the Ways & Means Committee summary. Giving NotebookLM “bite sized” items to analyze proved a much more accurate way to get it to focus on the proper items than trying to pick off multiple provisions at once. And I did scan the output to make sure I didn’t see any obvious issues, following up on a few things where I was surprised though most often discovering that “yeah, that’s how it was written” when I was done reviewing the appropriate sections of each document.

You can download that analysis below. There’s little reason for me to refine it more at this point since the final bill will require a “start over” (you never know what change will sneak in as a bill moves through the process) but, as I noted, with the caveat that you don’t assume it’s 100% accurate (which you shouldn’t do for any analysis–the law is what matters), I do find it useful.

Tax Provisions of The One, Big, Beautiful Bill, Title IX – Ways & Means Committee Tax ProvisionsDownload

If you haven’t used it, you might want to go look at NotebookLM. Even with a free account you can do a decent analysis (just not a huge number of them).