THE BIG FOUR’S CAMEL’S NOSE IS ABOUT TO GET IN THE LAWFIRM TENT

THE BIG FOUR’S CAMEL’S NOSE IS ABOUT TO GET IN THE LAWFIRM TENT

By Anthony J. Biller, Partner

KPMG, one of the Big Four accounting firms, has taken a significant step towards entering the US legal services market. On January 14, 2025, the Arizona Supreme Court’s Committee on Alternative Business Structures unanimously recommended that KPMG Law US, a subsidiary of KPMG, be granted a license to operate as an alternative business structure (ABS) in Arizona. If approved by the Arizona Supreme Court on January 28, this move would make KPMG the first Big Four firm to establish a law practice in the United States, potentially reshaping the legal services landscape. Arizona was the first state to allow nonlawyer ownership of law firms.

KPMG’s Proposed Legal Services Model

KPMG Law US aims to focus on large-scale, process-related legal tasks rather than high-stakes litigation or complex advisory work. The firm plans to offer services such as:

  • Synthesizing and re-drafting vendor contracts after significant M&A transactions
  • Volume contracting
  • Legal managed services
  • Contract lifecycle management

Christian Athanasoulas, a US tax practice leader at KPMG, is quoted as explaining their firm is not looking to compete in “bet-the-company” matters but rather to address areas where clients struggle with large-scale, process-related legal tasks.

Implications for the Legal Industry

KPMG’s potential entry into the US legal market has sparked discussions about:

1. Disruption of traditional legal practice structures: The move challenges the long-standing separation between legal and accounting services in the US.

2. Increased competition: Traditional law firms may face new competition in specific practice areas, particularly those involving process-driven work.

3. Multidisciplinary service offerings: KPMG’s model aims to integrate legal services with other consulting divisions, including tax and accounting.

4. Regulatory changes: Several states other than Arizona are already experimenting with pilot programs with alternative legal service providers or considering proposals allowing nonlawyers to provide specific legal services. KPMG’s application may inspire other states to consider similar regulatory shifts, potentially leading to broader changes in the legal services market.

The Alternative Business Structure (ABS) Model

Arizona’s ABS program, which began in 2021, allows nonlawyers to own or invest in law firms. This regulatory change has already permitted over 100 entities to receive approval, mainly in areas such as personal injury, mass torts, and trusts and estates law.

Looking Ahead

The Arizona Supreme Court’s decision on January 28, 2025, will be crucial in determining whether KPMG can proceed with its plans. If approved, it could pave the way for other Big Four firms and alternative legal service providers to enter the US legal market, potentially leading to significant changes in how legal services are delivered and consumed in the coming years.




The TRAIN ACT would be a Train Wreck for U.S. AI development.

By Anthony J. Biller, Partner

The Transparency and Responsibility for Artificial Intelligence Networks Act (TRAIN Act), introduced by Senator Welch on November 21, 2024, aims to create an administrative subpoena process for copyright owners to determine if their works were used in training AI models. The bill would allow copyright owners to request subpoenas from U.S. district courts, compelling AI model developers or deployers to disclose information about copyrighted works used in training their models.

This proposed legislation, while ostensibly aimed at protecting copyright holders, would likely prove detrimental for several reasons.

1. Undermining Fair Use: The use of copyrighted material for training AI models should be considered “fair use” under the Copyright Act. Fair use allows limited use of copyrighted material without permission from the copyright holder for purposes such as research, education, and innovation. AI training arguably falls under these categories, as it involves:

– Transformative use: AI training doesn’t reproduce works in their original form but uses them to create new, transformative outputs. The use is for the literal education of an advanced computing, non-human intelligence.

– Non-competitive purpose: Training data doesn’t compete with or replace the original works in the market.

– Potential for public benefit: AI models can lead to significant advancements in various fields, benefiting society at large.

As a “fair use,” there is no public policy justification for directing U.S. district courts to issue subpoenas for legal activities.

The justification of fair use for model training should not be conflated with fair use for model output, however. If an AI model produces or publishes content substantially similar to copyrighted content used in training, the model owners would and should be liable for copyright infringement, unless there is a legitimate defense of fair use independent of the fair use of educating the AI model.

2. Stifling Innovation: The bill would create an enormous administrative burden for U.S. AI developers, potentially slowing down research and development. This could lead to:

– Increased costs and time for AI development;

– Reluctance to use diverse training data, potentially reducing AI model quality and fairness; and

– A chilling effect on smaller companies and startups unable to bear the legal and administrative costs.

The ultimate purpose of copyright protection is to help encourage innovation and creativity by recognizing the rights of the content producer. The bill would allow copyright protection to be used in ways contrary to that purpose.

3. Global Competitive Disadvantage: The TRAIN Act would likely put the United States at a significant disadvantage in the global AI race:

– Other countries without such restrictions could develop AI more rapidly and efficiently.

– U.S. companies might relocate AI development to other countries with more favorable regulations.

– Foreign AI companies could gain a competitive edge, potentially dominating the global market.

4. Practical Challenges: The TRAIN Act also presents several practical challenges, both administratively and legally:

– Determining the exact copyrighted works used in training large AI models is often technically infeasible.

– The volume of potential subpoenas could overwhelm both the court system and AI companies.

– The rebuttable presumption of copying for non-compliance could lead to unfair legal outcomes.

5. Potential for Abuse: Copyright holders might use this process to harass AI companies or extract settlements, even in cases where fair use applies, which presumptively would be nearly all cases.

It warrants mentioning that effectively every person is a copyright “holder.” One does not need a federal copyright registration to “hold” a copyright. Every person who has ever posted a paragraph of content or an image they took to social medial is a copyright owner who might “believe” their content was used as part of training an AI model.

All that a copyright owner would need to invoke the subpoena power is a “subjective good faith belief” that the developer or deployer “used some or all” of copyrighted works to train an AI model. This “empty head but honest heart” minimal standard is, in effect, no standard at all for safeguarding the invocation of federal subpoena power.

If the AI developer or distributor does not timely respond, the copyright holder would be entitled to a presumption of copying and the full remedial provisions of the rules of civil procedure, which include being found in contempt of court and sanctioned.

In short, the TRAIN Act creates an invitation for abuse.

6. Privacy and Trade Secret Concerns: Forcing companies to disclose training data could force them to disclose valuable trade secrets.

Datasets used for training can be a key differentiator for AI companies. Disclosing this information could reveal valuable insights about a company’s AI strategy, potentially eroding their competitive advantage. It would also compromise the developers ability to claim their training models as trade secret.

*****

While the TRAIN Act aims to address legitimate concerns about copyright in the AI era, its approach is likely to cause more harm than good. It fails to recognize the transformative nature of AI training under fair use doctrine, potentially hampering U.S. innovation and competitiveness in this crucial field. Additionally, it presents several practical challenges and creates a legal environment ripe for abuse. A more balanced approach that considers both copyright protection and the unique challenges and opportunities of AI development would be more beneficial for all stakeholders and the United States’ position in the global AI landscape.




Alex Berenson v. President Joe Biden et al. (Amended Complaint)

Alex Berenson v. Joe Biden et al.




Understanding North Carolina Land Use Approvals

by Michael B. Kent, Jr.

More than 500 counties and municipalities in North Carolina have adopted some sort of zoning and land use ordinance. In general terms, these ordinances have two broad functions. First, they govern the uses to which land may be put by classifying individual parcels as residential, commercial, industrial, and so forth. Second, they govern the way land is developed by regulating the size and location of buildings, the amount of land that can be occupied, and other building and design issues.

These regulations can significantly affect real estate projects or transactions, and they frequently necessitate some type of government approval before a deal or project can be consummated. Failure to obtain the correct approvals or otherwise to comply with the local regulations could threaten the planned use of the property or expose the parties to fines, enforcement proceedings, and other legal problems.

To facilitate a better understanding of the regulatory system, some basic characteristics of land use approvals are highlighted below.

Types of Approvals

Land use approvals tend to fall into one of three basic categories—legislative, quasi-judicial, and administrative. These categories are named for the types of governmental decisions from which they result, and each type has its own distinctive characteristics, decisionmakers, and procedures.

Legislative Decisions

Legislative decisions adopt the ordinances, policies, and standards that make-up the regulatory system. These decisions are made by the locality’s governing board, such as the county commission or city council. In much the same way as Congress or the General Assembly enacts statutes, the governing board adopts ordinances and other legislative pronouncements that establish the rules governing land use and development.

The most typical approval resulting from a legislative decision is a rezoning. A rezoning occurs when the governing board adopts an ordinance that changes the zoning classification for a particular parcel (or parcels) of land. For example, a developer wanting to build a retail center on land zoned for residential use would need to have the governing board reclassify the land for commercial purposes. Because rezonings change the locality’s zoning map (an official map depicting the zoning classification for each parcel in the jurisdiction), they are also known as “map amendments.”

A rezoning usually requires at least two public hearings, one before the local planning commission (which recommends approval or disapproval) and the other before the governing board (which makes the final decision). Both hearings are open to the public, and citizen comments are allowed.

Quasi-Judicial Decisions

Quasi-judicial decisions apply the locality’s ordinances and policies to individual cases and situations. These decisions are often made by the locality’s board of adjustment, but other bodies (such as the governing board or planning commission) may also have the authority to make them in certain places or circumstances. These decisions derive their name from the fact that the party seeking the decision bears the burden of proving the existence of certain facts or the satisfaction of delineated legal criteria. In this way, quasi-judicial hearings can be thought of as “mini trials.” While such hearings are open to the public, only witnesses who can provide relevant evidence are allowed to speak.

The most frequently sought quasi-judicial approvals are special use permits and variances. Special use permits authorize a specific type of development or use within a designated zoning classification, but only if compliance with certain standards or conditions can first be shown. Local governments require these permits to mitigate potential harm from atypical (but perhaps not incompatible) uses. Places of worship or home occupations in a residential zone are common examples. Variances relax certain development standards that are proven to result in “unnecessary hardship” to the landowner. Essentially, this requires the landowner to demonstrate that something peculiar to the property makes strict application of the regulations inequitable or uniquely burdensome.

Administrative Decisions

Administrative decisions are those made by the locality’s planning and development staff. Legally, these decisions must be based on objective criteria prescribed in the regulations. The planning and development officials are not allowed to exercise subjective discretion but are charged simply with certifying whether a particular proposal meets the criteria or not. Common examples of administrative land use approvals include zoning compliance certificates, site plan approval, and subdivision approval. These approvals usually do not require a hearing, although one can be obtained by appealing an adverse decision to the local board of adjustment.

**********

Navigating the rules governing land use can be crucial to a successful real estate project or transaction. Envisage attorneys assist property owners, developers, purchasers, tenants, and other interested parties with investigating land use issues, developing strategies, and obtaining approvals. If you have questions about this alert or think we might be of assistance to you, you may contact us at (919) 755-1317.




Federal Trade Commission Votes to Ban Employment-Based Noncompetes

by Michael B. Kent, Jr.

On April 23, 2024, the Federal Trade Commission voted 3-2 to approve a final rule banning most noncompete provisions in the employment context. The rule has nationwide applicability and preempts inconsistent state and local laws. Additionally, the rule requires that employers notify employees subject to a prohibited noncompete that such provisions will not be enforced. If allowed to stand, the rule will change the landscape of employment law by overturning decades of state law precedents governing the validity and enforcement of noncompete agreements.

What is prohibited?

The rule deems each of the following acts to be “an unfair method of competition” and, thus, prohibited under federal law:

• To enter into a noncompete clause (or attempt to do so);
• To enforce a noncompete clause (or attempt to do so); and
• To represent that a worker is subject to a noncompete clause.

The rule defines “noncompete clause” broadly to include any term or condition of employment that prohibits, penalizes, or functions to prevent a worker from seeking or accepting work or operating a business in the United States. The rule prohibits only noncompete clauses that apply after the conclusion of employment, however. Restrictions imposed during the worker’s employment remain untouched.

The rule also contains a broad definition of “worker,” which encompasses not only traditional employees but also independent contractors, interns, apprentices, and volunteers.

Is the rule retroactive?

Not only does the rule prohibit an employer from entering into new noncompete clauses, but it also invalidates existing noncompetes. The rule does not apply, however, to causes of action related to a noncompete that accrued before the rule becomes effective.

What notices must an employer provide?

Employers must notify workers subject to existing noncompete clauses that enforcing the noncompete is now an unfair method of competition and that the clause cannot (and will not) be enforced against them. The notice can be hand-delivered, mailed, emailed, or texted to the worker, but it must be sent no later than the date on which the rule becomes effective. The rule contains model language for providing this notice, and employers who use the model language are deemed to have complied with the notice requirement.

Are there exceptions?

Aside from existing causes of action, the rule contains three exceptions:

Senior executives. Noncompetes entered into with senior executives prior to the rule’s effective date are not invalidated. New noncompetes may not be imposed after that date, however, even for senior executives. The rule defines “senior executive” as a worker who occupies a policy-making position and earns at least $151,164 in annual compensation.

Bona fide sale of business. The rule does not apply to noncompete clauses entered into as part of a bona fide sale of a business entity, of an ownership interest in a business entity, or of all or substantially all of a business entity’s operating assets. The rule defines “business entity” to include a partnership, corporation, association, limited liability company, or other legal entity, as well as divisions or subsidiaries of the foregoing.

Good faith. It is not a violation of the rule to enforce, attempt to enforce, or make representations about a noncompete clause where a good-faith basis exists to believe the rule does not apply to the noncompete in question.

When does the rule take effect?

By its terms, the rule becomes effective 120 days after its publication in the Federal Register. The FTC published the rule on May 7, 2024, meaning it becomes effective on September 4, 2024.

Is the rule legal?

To date, at least three legal challenges have been brought against the rule in federal courts. These lawsuits—two in Texas and one in Pennsylvania—allege several reasons for invalidating the rule, including a lack of authority on the part of the FTC, violations of separation of powers principles, and the irrationality of the rule’s economic impact and retroactive application. It remains to be seen whether these arguments will prove successful in delaying implementation of the rule or in striking it down altogether.

*****

Envisage attorneys assist both individuals and businesses with employment-related matters, and we are available to help you understand the new FTC rule and how it might affect you or your business. If you have questions about this alert or think we might be of assistance to you, you may contact us at (919) 755-1317.




Court Rules That COVID-19 Orders Violated Constitution

Court Rules That COVID-19 Orders Violated Constitution

by Michael B. Kent, Jr.

The North Carolina Court of Appeals recently issued an important decision addressing one of Governor Cooper’s COVID-19 “lockdown” orders. In North Carolina Bar & Tavern Association v. Cooper, the Court ruled that the order violated the North Carolina Constitution by allowing some businesses to reopen while forcing others to remain closed.

The case involved a challenge to Executive Order No. 141, which permitted certain drinking establishments—such as restaurant and hotel bars—to resume operations so long as they utilized certain safety precautions. The order required private bars to remain closed,
however, stating that “by their very nature,” such businesses “present greater risks of the spread of COVID-19.” Owners of the shuttered bars sued, alleging that the order effected a taking of their businesses and violated their constitutional rights.

Although the Court disagreed that the order constituted an unlawful taking, it found in favor of the bar owners on two of their other claims. First, the Court held that the order violated a provision of the state constitution guaranteeing to all North Carolinians “the enjoyment of the fruits of their own labor.” Second, the Court held that the order violated the state constitution’s guarantee of “the equal protection of the laws.”

The order ran afoul of both provisions, the Court ruled, because it resulted in an “irrational and arbitrary” distinction between those bars that could reopen and those that could not. The Court rejected the claim that the decision was based on “data and science.” Much of the government’s evidence came from the Governor’s “own assertions” in other executive orders or from ordinary news articles that could be found “by private citizens utilizing Internet search engines.” The one scientific study the government did produce post-dated Executive Order No. 141 (and, therefore, could not have been the basis for it) and failed to “differentiate between various types of bars.”

Because the evidence did not support the decision to allow some bars to reopen while forcing others to stay closed, the decision was “illogical and not rationally related” to the government’s goal of slowing the spread of the virus. That rendered the executive order unconstitutional, notwithstanding the public health concerns for which it was designed. “Our Constitution,” the Court explained, “applies even when a government official acts with the best stated purposes.”

Envisage attorneys help individuals and businesses protect their constitutional and other legal rights amidst increasingly complex regulatory landscapes (including a previous challenge to Executive Order No. 141). If you have questions about this alert or think we might be of assistance, you may contact us at (919) 755-1317.