Federal officials overseeing taxation, corporate financial reporting and public-company audits are increasingly using the term super intelligence (SI) for artificial intelligence, while exploring how the technology could change their work and the accounting profession. Their discussions reveal both an expanding interest in its capabilities and concern that regulatory oversight is not keeping pace with technological investment.
The shift in terminology follows a September executive order from President Donald Trump directing federal agencies to replace references to artificial intelligence and AI with super intelligence and SI in applicable government settings. At recent professional conferences, representatives of the Internal Revenue Service (IRS), Securities and Exchange Commission (SEC) and Public Company Accounting Oversight Board (PCAOB) have adopted that language while discussing practical applications, implementation risks and the need for stronger oversight. According to Accounting Today, the conversations also expose a broader challenge: regulators must understand how accounting organizations are deploying the technology before they can effectively supervise its use.
Washington changes terminology
The change in language has become increasingly visible in discussions involving the federal agencies responsible for taxation, securities regulation and public-company audit oversight. Officials from all three organizations have used SI during recent industry gatherings, reflecting the administration's preferred terminology rather than announcing a separate technological development or a new regulatory framework.
Trump signed an executive order in September 2026 titled Inaugurating the Era of Super Intelligence. The order established the administration's preference for the new terminology and directed executive-branch agencies, to the extent allowed by law, to use SI and super intelligence instead of the established AI terminology. It also called for federal officials not to recognize the older expressions in applicable settings.
The distinction matters when interpreting the regulators' statements. Their use of SI does not, by itself, establish that they are discussing a different class of technology from the artificial intelligence systems already used by businesses and government agencies. The terminology reflects the presidential directive. The substantive issues raised at the conferences concern familiar questions about data analysis, automation, productivity, financial accuracy and accountability.
The discussions took place at two prominent professional gatherings in Washington and the financial reporting community. At the American Institute of Certified Public Accountants' (AICPA) National Tax Conference, IRS officials described potential applications in tax administration. Later, representatives of the SEC and PCAOB discussed financial reporting, accounting-firm oversight and the modernization of audit inspections at Financial Executives International's Corporate Financial Reporting Insights conference.
Together, their comments show how the technology conversation is moving beyond general enthusiasm about automation. Regulators are considering where sophisticated data analysis could improve their operations, but they are also examining the consequences of allowing systems whose decisions may be difficult to explain to influence important financial and compliance processes.
IRS targets hidden noncompliance
The IRS sees potential value in SI primarily as a way to improve the identification of taxpayers and activities that deserve closer examination. Lia Colbert, commissioner of the agency's Small Business/Self-Employed division, discussed the possibilities at the AICPA National Tax Conference, describing the technology as a means of expanding what the agency can identify with its existing resources.
Her argument focused on the practical limits of manual enforcement. Tax authorities must examine information, identify patterns and decide where further investigative attention is justified. Better analytical capabilities could help the IRS distinguish higher-risk cases from routine activity, allowing employees to concentrate on complicated matters rather than spending as much time on preliminary administrative work.
Colbert described SI as a capacity-enhancing resource for the agency. The intended benefit is not simply processing more information. It is improving the selection of cases that deserve attention, particularly where patterns of potential noncompliance might otherwise remain difficult to detect.
She said the IRS expects the technology to help reveal circumstances that earlier analytical approaches might have struggled to identify. That could improve the agency's ability to recognize areas of tax noncompliance and direct its available personnel toward cases that require more substantial examination.
The IRS specifically expects SI to strengthen its efforts to identify people who fail to submit required tax returns. Colbert said the agency wants to use the technology to select more valuable enforcement work and make better decisions about where its investigative capacity should be deployed.
The nonfiler strategy is an established part of tax administration. People who are required to file returns but do not submit them can represent a compliance problem for the government. Identifying those individuals and encouraging or requiring them to meet their obligations involves more than processing returns that have already arrived. It also requires finding cases where expected filing activity is absent.
Colbert indicated that SI could make this work more effective by helping the agency identify relevant cases and prioritize its response. Rather than treating every potential nonfiler as an equally important investigative target, the IRS wants to identify situations that merit more attention because of their potential significance.
The same approach could extend to another enforcement area: benefits fraud and abuse. Colbert said the IRS had already been speaking publicly about these concerns during the previous fiscal year and suggested that analytical technology could allow the agency to investigate the issue more deeply.
She connected the two areas through a common operational objective. Whether the agency is identifying missing tax returns or examining possible misuse of benefits, improved analytical capabilities could reduce the amount of manual work needed to locate and assess relevant cases.
However, her remarks described anticipated uses rather than a completed nationwide deployment. She did not provide specific performance figures, an implementation timetable or a quantified estimate of how much enforcement activity SI might generate. The emphasis was on the opportunities the technology could create for more selective use of agency resources.
Taxpayer Advocate sees better filters
Erin Collins, the National Taxpayer Advocate, also discussed SI at the AICPA conference. Her comments approached the subject from a different perspective, focusing on the possibility that improved data analysis could help the IRS distinguish genuine taxpayer activity from misleading information generated by changes in personal circumstances.
Collins described a hypothetical situation involving a taxpayer whose address and employer change after moving to Alaska. A conventional screening process might flag the return because several pieces of information differ from what the IRS previously held. Such a change could trigger a filter even if the taxpayer has done nothing wrong.
The challenge is determining whether the change represents legitimate activity or a potential compliance issue. A screening system that notices differences can identify cases for further attention, but the initial alert does not necessarily establish that the information is inaccurate or that the taxpayer has acted improperly.
Collins suggested that additional data sources could help establish whether the individual had genuinely moved and purchased a home in Alaska. By examining information beyond the details that triggered the initial filter, the IRS might be able to verify the taxpayer's circumstances more effectively.
Her example highlights a possible improvement in the relationship between automated screening and subsequent verification. A filter can identify an unusual combination of details, but additional analytical capabilities could help establish whether those details make sense when considered alongside other available information.
This could matter for taxpayers because incorrect or incomplete interpretations of changed information can create unnecessary compliance complications. The example was hypothetical, however, and Collins did not claim that the IRS had already implemented a system capable of resolving such cases automatically.
She described the agency as working through how to accelerate the initial stage of its screening process. SI could potentially help bring together additional information, assess whether a flag deserves further investigation and improve the way the IRS evaluates cases after an initial filter has identified them.
The two IRS perspectives complement one another. Colbert concentrated on identifying higher-value enforcement opportunities and reducing manual work. Collins focused on whether additional information could make screening and verification more accurate. Both see potential in using technology to make better use of information already relevant to tax administration.
Neither official presented SI as a substitute for the agency's underlying compliance responsibilities. Instead, their comments suggest that the IRS is interested in using improved analytical capabilities to make existing processes more targeted and efficient.
SEC examines financial reporting
The SEC's concerns are different because its focus is not primarily tax enforcement. The commission oversees securities markets and public-company reporting, making the accuracy and reliability of corporate financial information central to its responsibilities.
Kurt Hohl, the SEC's chief accountant, discussed the agency's work with companies, auditors and other regulators at Financial Executives International's Corporate Financial Reporting Insights conference. He said the commission had conducted outreach meetings with industry participants to understand how SI was being used, or not used, in financial reporting and internal control over financial reporting.
Internal control over financial reporting, commonly abbreviated as ICFR, refers to the processes and safeguards companies use to support accurate financial records and reliable financial statements. These controls help reduce the risk that errors or inappropriate transactions will undermine the information investors receive.
The SEC's outreach also includes engagement with other regulatory bodies, particularly the PCAOB. Hohl said the purpose is to understand the changing technological landscape and determine where additional guidance might be useful to encourage responsible adoption.
The commission sees potential benefits in innovation that supports dependable financial reporting and effective internal controls. But Hohl's comments made clear that enthusiasm for the technology does not eliminate the requirements companies already face.
He observed that organizations appeared to be approaching SI deployment more cautiously in financial reporting than in some operational functions during the later months of the year. That distinction is important because companies can use productivity-oriented tools for many routine activities without necessarily allowing them to influence the preparation of formal financial information.
Some AI-powered productivity tools have already become important to ordinary business operations, according to Hohl. Yet applying similar technologies to accounting records, financial reporting and control processes introduces questions that require more careful examination.
A system that helps employees complete routine administrative tasks does not necessarily present the same risk as one that influences accounting judgments or the information supporting financial statements. The consequences of errors can be particularly significant when technology becomes part of a process responsible for producing information relied upon by investors and regulators.
Hohl emphasized that the existing requirements remain in place. Companies must maintain accurate records, effective accounting controls, appropriate risk assessments and reliable financial statements regardless of whether SI is involved in their preparation.
The SEC's central message is that adopting SI does not transfer or reduce a company's responsibility for accurate financial reporting. Technological innovation must operate within the established framework of accounting accountability and internal control.
Three priorities for companies
Hohl identified three areas that should guide organizations considering SI deployment: governance, risk assessment and effective responses through internal controls. His recommendations are intended to help businesses understand the technology before allowing it to become embedded in important financial processes.
The first priority is governance. Hohl said companies should have a defined strategy for adopting SI rather than allowing individual employees or departments to introduce tools without an overarching organizational plan.
That strategy should address the reliability of the data used by the technology and the preparation of employees who will work with it. Companies also need to consider how the technology fits into their broader business processes and how responsibility for its use is distributed throughout the organization.
The role of company boards is another important element. Hohl specifically highlighted audit committees, which oversee aspects of financial reporting and the external audit relationship. Directors should understand the organization's SI strategy, the risks associated with deployment and the measures management intends to use to address those risks.
The second priority is effective risk assessment. Hohl argued that companies should examine individual processes to determine exactly where SI is being used and how its involvement changes the risks associated with those processes.
That assessment cannot stop at identifying which software a company has purchased. Organizations need to understand the functions performed by the technology, the information it processes and the potential consequences of its outputs being inaccurate or unsuitable.
Hohl highlighted several characteristics that distinguish these systems from more conventional business technologies. These include opacity, probabilistic behavior and adaptivity. Together, these characteristics can make it more difficult to understand precisely how a system reaches a particular result and whether its behavior will remain consistent across different situations.
Opacity refers to the difficulty of examining or explaining the internal reasoning behind certain system outputs. Probabilistic behavior means that results can be based on likelihoods rather than a guaranteed identical response to the same type of input. Adaptivity describes the ability of systems to change their behavior in response to data or changing conditions.
Hohl said companies need to understand these characteristics when evaluating the risks introduced into financial reporting. The relevant question is not simply whether the technology works in a general sense, but whether its use is appropriate and sufficiently controlled in the particular process where it has been introduced.
He also warned that businesses must look beyond systems they operate directly. SI may enter financial reporting processes through external service providers, including organizations responsible for performing outsourced functions.
That creates a potentially complicated oversight environment. A company might rely on an outside provider for a particular process while remaining dependent on technology used by that provider. Even when the software and operational procedures are outside the company's direct control, their outputs can still affect its financial reporting.
Hohl stressed that outsourcing does not eliminate the company's responsibility for the process. Organizations therefore need to understand how third parties use SI and what safeguards exist around the technology, particularly when outsourced services contribute to financial information.
The third priority follows from the first two: companies need effective internal controls that respond to the particular risks created by SI. Governance identifies the strategy and responsibilities. Risk assessment establishes where problems could emerge. Internal controls provide the safeguards intended to prevent, identify or address those problems.
The three priorities form a connected approach rather than separate administrative exercises. A company cannot establish meaningful controls without first understanding where the technology operates and what risks it introduces. Likewise, a governance strategy that ignores third-party systems may leave important parts of the financial reporting process outside effective oversight.
PCAOB wants modern inspections
The PCAOB faces a different technological challenge. Rather than concentrating only on how businesses use SI in financial reporting, the organization must also consider how to modernize its own oversight of accounting firms.
Demetrios Jim Logothetis, the PCAOB chairman, discussed the issue at the Financial Executives International conference. His remarks highlighted a gap between the technological investments being planned by accounting firms and the tools available to the regulator responsible for inspecting their audit work.
The PCAOB oversees audits of public companies and registered accounting firms. Its inspection responsibilities are intended to assess the quality of audit work and identify deficiencies that could affect the reliability of financial reporting.
Logothetis acknowledged that the regulator is behind the accounting industry in technological terms. He said firms were preparing substantial investments and developing plans involving technology, even though SI had not yet become broadly established throughout the profession.
His concern was that the PCAOB needed to develop its own capabilities rather than allow the firms it supervises to move substantially ahead of its inspection infrastructure.
The chairman said the organization was looking for an overall modernization plan. That plan should cover software and technology, potentially including hardware, but he placed particular importance on identifying the resources required to support a more technologically capable inspection process.
The scale of the challenge becomes clearer when considering the PCAOB's existing working methods. Logothetis said inspectors have relied mainly on Microsoft Excel spreadsheets and Word documents to carry out much of their work.
These familiar office applications can support documentation and analysis, but the chairman wants the inspection process to become more efficient and effective through broader use of available technology.
One possibility he described involves working with accounting firms to obtain information digitally and use technological tools to improve the inspection process. The objective is to move away from an approach that depends heavily on conventional documents and manual handling of information.
This does not necessarily mean that SI itself will perform inspections or make regulatory decisions. Logothetis discussed a broader modernization effort involving technology and resources, with the goal of making oversight more capable of dealing with the increasingly sophisticated systems used by the profession.
The PCAOB's position also illustrates why regulators cannot focus exclusively on the technology used by the organizations they supervise. If accounting firms change how they collect information, conduct audits and evaluate financial statements, inspectors need appropriate tools and expertise to understand those changes.
Without that capability, the regulator could struggle to assess the effectiveness of new processes or determine whether firms have maintained appropriate standards as their technology investments expand.
Human oversight remains essential
Despite his interest in modernization, Logothetis warned against assuming that technology will automatically improve audit quality. He urged audit committees to examine how both public companies and their accounting firms are using AI, including the consequences for audit scope, fees and commitments.
His concern is that expectations about the technology could move faster than organizations' ability to demonstrate what it actually delivers. Firms may make substantial investments and promises, but those commitments need to be evaluated against the work performed and the results achieved.
Logothetis said companies and auditors should take a measured approach rather than allowing excitement about technological possibilities to determine professional decisions. Understanding the practical implications of deployment is particularly important when the technology affects the work performed during an audit.
For audit committees, this means examining more than whether an accounting firm has adopted sophisticated software. They should understand how the technology changes the scope of the engagement, whether it affects fees and what commitments the firm is making about its capabilities.
The chairman also identified the combined effect of technology, organizational scale and emerging issues as a source of risk. Large accounting organizations may be investing heavily in new systems, while regulators must understand how those systems are being integrated into professional processes.
His remarks placed particular emphasis on maintaining human involvement. Technology can improve processes, but its outputs should not be accepted as conclusive simply because they were produced by an advanced system.
"We are overseeing the accounting firms, and yet we're way behind"
— Demetrios Jim Logothetis, PCAOB chairman
The warning is especially relevant when technology is used in audit work, where professional judgment and evidence evaluation remain central. Automated analysis can support an auditor's work, but the presence of a technological result does not establish that the result is correct or that the underlying evidence has been adequately examined.
Logothetis argued that regulators need to understand what accounting firms are actually doing and how they can provide useful oversight. That requires keeping expectations grounded in demonstrated effectiveness rather than allowing technological promises to replace proper evaluation.
His comments also complement Hohl's position at the SEC. Both officials recognize the potential benefits of SI while emphasizing that companies and accounting firms must retain responsibility for the quality of their financial reporting and audit work.
The difference is primarily institutional. Hohl concentrated on how organizations should govern and control the technology within financial reporting. Logothetis focused on the PCAOB's ability to inspect firms that are making technological investments and the need for human judgment throughout the process.
Shelved PCAOB projects return
Logothetis also addressed questions from Accounting Today about whether the PCAOB might reconsider projects that were put aside during the previous year, before he became chairman. The discussion covered two initiatives: NOCLAR and firm and engagement metrics.
NOCLAR is an abbreviation for non-compliance with laws and regulations. In the accounting profession, the subject concerns how auditors and other relevant parties should address indications that an organization has failed to comply with applicable legal or regulatory requirements.
Firm and engagement metrics concern the measurement and reporting of information about accounting firms and individual audit engagements. Such measures can provide stakeholders with information that helps them assess aspects of audit performance and the organizations conducting the work.
Both initiatives had attracted attention from stakeholders, but Logothetis said they were also potentially controversial. He explained that he had been largely outside the profession when the original discussions took place, having retired and not followed the projects closely at the time.
His return to the PCAOB has given him a different perspective. After hearing from stakeholders, he considers both subjects important enough to warrant further examination.
However, he did not commit to restoring either project in its previous form. Instead, he indicated that the board would approach the issues afresh, considering their advantages, disadvantages and the concerns raised by interested parties.
That distinction leaves the eventual outcome open. A decision to reconsider a project does not mean that its original proposals will automatically return or that the regulator has already settled on a new approach.
Logothetis said both matters would remain on the PCAOB's agenda for appropriate consideration. The organization intends to assess them based on the issues involved and the perspectives of stakeholders rather than treating earlier decisions as binding conclusions.
His comments suggest that the new leadership is willing to revisit subjects that matter to parts of the profession while avoiding an advance commitment to a particular regulatory outcome.
Why it matters
The conversations at the AICPA and Financial Executives International conferences reveal a common challenge across three federal oversight organizations: technology is creating opportunities to process information differently, but the institutions responsible for tax compliance, financial reporting and audit quality must establish how those capabilities can be used responsibly.
For the IRS, the immediate attraction is operational. Better analytical tools could help identify nonfilers, select higher-value compliance cases and examine possible benefits fraud. Colbert's comments point toward a more targeted enforcement model, while Collins highlighted the potential to improve the accuracy of taxpayer screening through additional information.
For the SEC, the priority is preserving the reliability of corporate financial information. Hohl's three-part framework places responsibility on companies to establish governance, understand process-level risks and maintain controls that account for the distinctive behavior of SI systems. The same responsibilities extend to technology supplied by outside service organizations.
For the PCAOB, the issue is whether audit oversight can keep up with the firms being inspected. Logothetis wants a modernization strategy that gives inspectors better technological resources and reduces reliance on traditional document-based processes. At the same time, he wants audit committees and regulators to challenge claims about technology rather than accept them without evidence.
The administration's terminology change provides the political backdrop, but the regulators' practical concerns extend well beyond vocabulary. Their remarks show that the adoption of SI is being considered within existing institutional responsibilities, including tax enforcement, accurate recordkeeping, corporate accountability and independent audit oversight.
Several developments will help clarify how far these ambitions progress. For the IRS, the key issue is whether its proposed applications translate into more effective identification of nonfilers and potential benefits abuse. For the SEC, continued engagement with companies, auditors and the PCAOB will help establish whether additional guidance is needed as organizations introduce SI into financial reporting.
The PCAOB's modernization plans will also deserve attention. Its ability to develop a broader technological inspection process could influence how effectively it evaluates accounting firms as their investments grow. The regulator's review of NOCLAR and firm and engagement metrics will provide another indication of how the new leadership intends to approach unresolved oversight questions.
For accounting firms and public companies, the immediate message is less ambiguous. Technology investments do not replace professional accountability, and outsourcing does not transfer responsibility for the quality of financial processes. Organizations will need to understand where their systems operate, how their outputs affect important decisions and whether human reviewers have the information and authority necessary to challenge them.
The central regulatory challenge is no longer simply whether accounting organizations will adopt SI, but whether their controls, inspectors and professional judgment can keep pace with its use. The IRS, SEC and PCAOB are signaling that they see potential benefits, but none of their remarks suggests that technological adoption should come at the expense of reliable evidence, effective oversight or human responsibility.