How Is This Bill Enforced
Verbatim statutory text on the left; plain-language analysis and a per-section checklist on the right. Numbered markers cross-link to the matching checklist row.
This Act may be cited as the ''Research and Oversight of AI in Courts Act of 2026''.
Establishes the short title of the Act as the Research and Oversight of AI in Courts Act of 2026. No compliance obligations are created.
(a) The purpose of this Act is to provide Congress with— (1) an analysis of any issues or benefits related to the use of AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) and automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3); (2) a review of the use of AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) and automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3); and (3) any recommendations based on the analysis and review conducted under paragraphs (1) and (2).
States the purpose of the Act: to provide Congress with an analysis of issues and benefits related to AI speech-to-text and automatic speech recognition technology in courts, a review of current use, and recommendations. This is a legislative-purpose statement and creates no compliance obligations.
(b) 1 Not later than 60 days after the date of enactment of this Act, the Attorney General of the United States, acting through the Director of the National Institute of Justice, shall establish a task force (to be known as the ''AI Research and Oversight in Courts Task Force'') to assess the feasibility, accuracy, privacy, and civil liberty implications of using AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) and automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3).
Directs the Attorney General, acting through the Director of the National Institute of Justice, to establish the AI Research and Oversight in Courts Task Force within 60 days of enactment. The task force's mandate is to assess the feasibility, accuracy, privacy, and civil liberty implications of AI speech-to-text and automatic speech recognition technology in the judicial system. This is a governmental organizational mandate — it imposes an obligation on the executive branch but not on private parties.
(c)(1)–(2) The duties of the task force established under subsection (b) shall be to— (1) assess policy findings, regulatory findings, and legal findings with respect to the use of AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) and automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3); and (2) make recommendations to Congress, the executive branch, and the judicial branch for any judicial, legislative, or regulatory reforms to ensure the use of AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) and automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) does not infringe upon any rights under the Constitution of the United States, including the litigant's right to an accurate official court record.
Defines the two core duties of the task force: (1) assess policy, regulatory, and legal findings regarding AI speech-to-text and automatic speech recognition technology in courts; and (2) make recommendations to Congress, the executive branch, and the judicial branch for reforms to ensure these technologies do not infringe constitutional rights, including the right to an accurate official court record. These are internal governmental study mandates, not compliance obligations on regulated parties.
(d)(1) The task force established under subsection (b) shall be composed of 15 members, to be appointed by the Director of the National Institute of Justice, as follows: (A) 4 members employed by the Federal Government in at least one of the following capacities: (i) As an employee of the National Institute of Justice. (ii) As an employee of the Administrative Office of the United States Courts. (iii) As a clerk of court. (iv) As a Federal judge. (v) As a Federal prosecutor. (B) 11 members not employed by the Federal Government, which shall include the following individuals: (i) An individual who is a member of a professional association that specializes in the official record making process of the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) and in the technology used for such record keeping. (ii) An individual who is a specialist in civil liberty law. (iii) An active State court judge or a retired Federal court judge with experience reversing a ruling because of a written or audible deficiency in the official court record.
(d)(2) A member appointed under paragraph (1)(B) to the task force established under subsection (b), may not be employed by, contracted with, receive compensation from, or otherwise represent any entity that develops, markets, sells, or provides artificial intelligence technologies or related services.
(d)(3) Members appointed to the task force established under subsection (b) under paragraph (1) of this subsection shall have expertise in— (A) the official record making process in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3); or (B) the technology used by the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) to create and modify a court record.
(d)(4)–(6) CO-CHAIRS.—The Director of the National Institute of Justice shall designate as co-chairs of the task force established under subsection (b)— (A) 1 member appointed to the task force under paragraph (1)(A); and (B) 1 member appointed to the task force under paragraph (1)(B). (5) VACANCIES.—In the case of a vacancy on the task force established under subsection (b), the co-chairs appointed under paragraph (4) shall appoint a new member not later than 15 days after the vacancy occurs. (6) COMPENSATION.— (A) FEDERAL EMPLOYEES.—Each member of the task force established under subsection (b) appointed under paragraph (1)(A) of this subsection shall serve without compensation in addition to that received for services as an employee of the United States. (B) OTHER MEMBERS.—Each member appointed to the task force established under subsection (b) under paragraph (1)(B) of this subsection shall serve without compensation. (C) REIMBURSEMENT.—Each member appointed to the task force established under subsection (b) under paragraph (1) of this subsection shall be allowed travel expenses, including per diem in lieu of subsistence, at rates authorized for employees of agencies under subchapter I of chapter 57 of title 5, United States Code, while away from their homes or regular places of business while carrying out the duties described under subsection (c).
Specifies the composition, eligibility, expertise requirements, co-chair designation, vacancy-filling procedures, and compensation rules for the 15-member task force. Four members must be federal government employees (from the National Institute of Justice, Administrative Office of the U.S. Courts, clerks of court, federal judges, or federal prosecutors). Eleven must be non-federal members, including a member of a professional association specializing in court record-making, a civil liberty specialist, and a state or retired federal judge with experience reversing a ruling due to court record deficiency. Notably, non-federal appointees may not be employed by, contracted with, or receive compensation from any entity that develops, markets, sells, or provides AI technologies. All members serve without additional compensation beyond travel reimbursement.
(e) 2 Not later than 18 months after the task force is established under subsection (b), the task force shall submit to the Attorney General, the Committee on the Judiciary of the Senate, and the Committee on the Judiciary of the House of Representatives a final report on the following: (1) Whether a court record created or modified by AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) affects the quality or accuracy of a transcription of a legal proceeding. (2) Whether AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) alters any comments made in a courtroom by an individual with a speech impediment, unique speech pattern, accent, or dialect. (3) Whether using AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) to perform tasks in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) will increase or decrease costs for litigants. (4) An assessment of any cybersecurity risks associated with using AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) to perform tasks in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3). (5) An assessment of any other risks associated with using AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) to analyze or process a court record and perform tasks in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3). (6) An assessment of any effects on the data integrity, authenticity, or preservation of the evidentiary value of a court record when using AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) to manage, process, and store such court record. (7) Whether implementing AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) to conduct managerial and administrative tasks in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) will increase or decrease court expenditures. (8) Whether a court record created or modified by AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) should display a watermark, header, footer, or banner to indicate that such court record was created or modified by such technology. (9) Whether a watermark, header, footer, or banner described under paragraph (8) should be displayed permanently on a court record. (10) Whether metadata should be included in a court record created or modified by AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2), including— (A) which AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) tool was used to create or modify such record; (B) which version of the AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) was used to create or modify such record; and (C) any changes made by such technology to such record. (11) Whether the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) should provide guidance on the selection process of vendors providing AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) to ensure the safety and privacy of parties and individuals involved in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3). (12) Whether the use of AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) compromises the accuracy or integrity of a court record. (13) Whether the use of AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) or automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) causes any disruptions during a judicial proceeding. (14) An analysis of any potential future developments and advancements of AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) and automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2) within 10 years of the date of enactment of this Act, and how the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) should adjust its policies to keep pace with the quickly evolving technology. (15) Any recommendations for policies to be implemented in the United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3) with respect to AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1) and automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2).
Requires the task force to submit a final report to the Attorney General and the Judiciary Committees of both chambers within 18 months of establishment. The report must address 15 enumerated topics covering transcription accuracy, speech impediment and dialect effects, cost impacts for litigants, cybersecurity risks, data integrity and evidentiary value, court expenditures, watermarking and metadata for AI-modified records, vendor selection guidance, accuracy and integrity concerns, disruptions during proceedings, future technology developments, and policy recommendations. This is a government-to-government reporting obligation and does not impose compliance duties on private parties.
(f) Not later than 4 months after the date of enactment of this Act, and every 4 months thereafter until the task force established under subsection (b) submits the final report under subsection (e), the task force shall submit to the Committee on the Judiciary of the Senate and the Committee on the Judiciary of the House of Representatives a report on— (1) the status of the final report; and (2) whether the final report will be submitted to such committees by the deadline established under subsection (e).
Requires the task force to submit status reports to the Judiciary Committees every four months, beginning four months after enactment, on the progress of the final report and whether it will meet the 18-month deadline. This is a government-internal procedural requirement.
(g) The task force established under subsection (b) shall terminate, and this Act shall cease to be effective, upon the date of submission of the final report under subsection (e).
Provides that the task force terminates, and the Act ceases to be effective, upon submission of the final report. This is a sunset provision creating no compliance obligation.
(h)(1) The term ''AI speech-to-text technologyAI speech-to-text technologyThe term "AI speech-to-text technology" means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.Sec. 2(h)(1)'' means a machine-based system that uses artificial intelligence, including machine learning and natural language processing, to automatically perceive, analyze, and convert human speech into written text or machine-readable data, enabling real-time transcription, accessibility, and interaction with digital systems.
(h)(2) The term ''automatic speech recognition technologyAutomatic speech recognition technologyThe term "automatic speech recognition technology" means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.Sec. 2(h)(2)'' means a technology that processes human speech and converts it into a written or machine-readable format without using artificial intelligence to facilitate communication and accessibility.
(h)(3) The term ''United States judicial systemUnited States judicial systemThe term "United States judicial system" means all State and Federal courts in the United States and the territories of the United States.Sec. 2(h)(3)'' means all State and Federal courts in the United States and the territories of the United States.
Defines three key terms used throughout the Act: AI speech-to-text technology (AI-powered systems that convert speech to text), automatic speech recognition technology (non-AI systems that convert speech to text), and United States judicial system (all state and federal courts, including territorial courts). The distinction between the first two terms is notable — the bill scopes both AI-based and non-AI-based transcription technologies, treating them as separate categories subject to the same study mandate.