State courts are finding new ways to serve the public — benefiting both court users and staff. This report explores how artificial intelligence is showing promise in helping courts improve services, especially for self-represented litigants and in monitoring guardianships and conservatorships. In addition to artificial intelligence (AI), it explores other topics shaping the future of court systems, including innovations in jury participation, the use of guided interviews to assist litigants, how courts are addressing ongoing challenges in housing courts, and ways to use data storytelling for effective communication. The publication also offers practical advice, such as 10 steps to launch an AI project and steps to develop a judicial learning center. These insights can help courts navigate the evolving landscape of court services and technology. As AI continues to evolve, its application in legal aid organizations presents a unique opportunity to address the access to justice crisis by centering on the needs of their human constituents while minimizing administrative work, reallocating talent, and empowering clients.

This Council for State Governments Justice Center analysis shows that after earlier declines in prison admissions across the country, states are now on sharply different trajectories—some continuing to lower admissions while others reverse course. These shifts underscore the need for state leaders to understand how their supervision systems are shaping prison populations, particularly when community supervision’s proportion of the prison system made up 40% of all state prison admissions in 2023. Since 2018, four distinct trends in prison admissions overall (that is, all admissions and not just admissions related to supervision violations) have emerged. In 2 states (Arkansas and Tennessee) have seen significant spikes in prison admissions, in 7 states (including Florida) admissions returned to near-2018 levels. Although admissions in 25 states have started increasing since the lowest levels seen in 2021, they have not returned to the higher levels seen in 2018. In 14 states, prison admissions dropped sharply in 2021 and have remained low since. As of 2023, they were still 30% to 62% below 2018 levels. In 2023, nearly 200,000 people were admitted to prison for violating probation or parole, including over 110,000 people for technical violations. Despite some progress since 2018, supervision continues to function more as a pathway into prison than a support system for success.  Notably, some states have achieved real, lasting progress. For example, Georgia saw a 25% drop in technical violation admissions. In contrast to public perception, supervision violations for new criminal activity accounted for less than 2% of all arrests in 2023. Only 5% of people on parole (0.5% of total arrests) were returned to prison for a new crime violation. Fewer than 2% of people on probation (1% of total arrests) were incarcerated for new offense violations. 

The Illinois Department of Transportation provides Sustained Traffic Enforcement Program (STEP) grants to local and state agencies to operate High Visibility Enforcement (HVE), a universal traffic safety approach designed to create deterrence and to change unlawful and dangerous driving behaviors through the use highly visible and proactive law enforcement, and increase campaigns targeting impaired driving and occupant protection violations during holiday weeks. The program aims to reduce fatalities and severe injuries due to motor vehicle crashes. This study investigates the effectiveness of STEP on road traffic safety. Using panel data modelsit analyzes how and whether STEP participation status and grants reduce total crash counts, the share of fatal and severe injury crashes, and the share of alcohol-impaired crashes. The counties that have STEP participating agencies had lower rates of fatal, severe injury, and impaired driving crashes the following year compared to counties without participating agencies. Further, the grant dollars were associated with a reduction in the share of fatal and severe injury crashes in the grant-awarded year. Recommendations include using HVE to discourage dangerous driving behaviors during holidays and special events and conducting thorough evaluations on enforcement performance to increase effectiveness.

This data report includes nine sortable tables that show how each of the states and the District of Columbia compare with one another and with the nation on demographics, state residents’ highest level of education, faculty pay, college enrollment, diversity, graduation rates, tuition costs, state aid, and more. In Florida, 95% of students take the SAT with an average score of 948 (the average SAT score nationally was 1024). Fewer Florida students, 44%, take the ACT; with an average score of 19.0 (the average ACT score nationally was 19.4). Florida’s six-year graduation rate for first-time, full-time, degree-seeking students who entered degree-granting four-year institutions in the fall of 2017 and graduated within six years was 67.8% compared to a national six-year graduation rate of 64.9%.

Curricular analytics – systematic analysis of curricula data to inform program and course refinement – becomes an increasingly valuable tool to help institutions align academic offerings with evolving societal and economic demands. Large language models (LLMs), which are advanced AI systems that understand and generate natural language, or human-like text, are promising for handling large-scale, unstructured curriculum data, but it remains uncertain how reliably LLMs can perform curricular analytics tasks. In this paper, the researchers systematically evaluate four text alignment strategies based on LLMs or traditional Natural Language Processing (NLP) methods for skill extraction, a core task in curricular analytics. Using a stratified sample of 400 curriculum documents of different types and a human-LLM collaborative evaluation framework, the authors find that retrieval-augmented generation is the top-performing strategy across all types of curriculum documents, while zero-shot prompting performs worse than traditional NLP methods in most cases. The findings highlight the promise of LLMs in analyzing brief and abstract curriculum documents, but also reveal that their performance can vary significantly depending on model selection and prompting strategies. This underscores the importance of carefully evaluating the performance of LLM-based strategies before large-scale deployment.

For years, college and career readiness has been the mantra of many education leaders. States adopted higher standards, aligning expectations in K–12 schools with what their public colleges require incoming first-year students to know. States developed new “portraits of a graduate” to articulate the skills, competencies, and knowledge today’s students need to thrive. States broke down silos and linked their education and workforce data to better understand students’ pathways to postsecondary education and careers—and where they faced roadblocks and barriers. The data allows states to begin not only measuring whether their high school students graduate and are ready for college or a career but also holding their high schools accountable for it. Forty-two states, including Florida, currently use at least one college and career readiness indicator for federal or state high school accountability requirements, and 16 states have multiple indicators. Although there are trends across states, state leaders continue to make distinct choices, based on their own priorities and goals, about how to design and deploy these indicators. For example, 36 states, including Florida, design its indicators so that all the measures in the indicator are interchangeable. In addition, 39 of the 42 states include both college and career readiness measures, and 20 of these states also measure military or civic readiness. Advanced Placement and International Baccalaureate coursework and exams are the most common college readiness measures (35 states), followed by dual or concurrent enrollment (34 states). There is more variance in career-ready measures, but the most popular option is student attainment of industry-recognized credentials (23 states), despite evidence that the value of many of these credentials in the labor market might be limited.

This report provides monthly and episodic poverty estimates for 2023, using data from the 2024 Survey of Income and Program Participation. The survey is a nationally representative, longitudinal survey administered by the U.S. Census Bureau that provides comprehensive information on the dynamics of income, employment, household composition, and government program participation. Highlights from the report include that the monthly poverty rate was 12.9% in January, and it did not significantly change throughout 2023. More than 1 in 7 individuals (15.7%) were in poverty for at least 2 consecutive months and 21.6% of non-Hispanic Black individuals and 23.0% of Hispanic individuals were in poverty for at least 2 consecutive months. Of those who experienced episodic poverty, 60.4% were in poverty all 12 months of 2023.

Recent growth in artificial intelligence (AI) capabilities has spurred a corresponding rise in public interest. Developments in generative AI— technology that can create text, images, audio, video, and other content when prompted by a user—have revolutionized its applications in various industries. The U.S. Government Accountability Office (GAO) was asked to describe federal agencies’ efforts to pursue generative AI. This report is the fourth in a body of work on generative AI. The GAO’s objectives included describing the ongoing and planned uses of generative AI by selected agencies, as well as the resulting potential benefits. Across the selected agencies the GAO reviewed with artificial intelligence (AI) inventories, the total number of reported AI use cases nearly doubled from 571 in 2023 to 1,110 in 2024. At the same time, generative AI use cases increased about nine-fold, from 32 to 282. The GAO also found that generative AI offers potential benefits. In the mission-support area, the technology could improve written communications, information access efficiency, and program status tracking. Program-specific examples include the U.S. Department of Veterans Affairs using generative AI to automate various medical imaging processes to enhance veterans’ diagnostic services, and the U.S. Department of Health and Human Services using generative AI to extract information from publications and identify outbreaks in areas previously thought to be polio-free to support containment of the poliovirus. However, agency officials report that they face several challenges to using generative AI, such as complying with existing federal policies and guidance, having sufficient technical resources and budget, and maintaining up-to-date, appropriate use policies. Agencies are beginning to address these challenges by (1) leveraging available AI frameworks and guidance to inform their policies and (2) engaging in collaborative efforts with other agencies.

Data centers are among the fastest-growing electricity consumers, raising concerns about their impact on grid operations and decarbonization goals. Their temporal flexibility—the ability to shift workloads over time—offers a source of demand-side flexibility. This paper models power systems in three U.S. regions: Mid-Atlantic, Texas, and the Western Interconnect (WECC), under varying flexibility levels. This paper evaluates flexibility's effects on grid operations, investment, system costs, and emissions. Across all scenarios, flexible data centers reduce costs by shifting load from peak to off-peak hours, flattening net demand, and supporting renewable and baseload resources. This load shifting facilitates renewable integration while improving the utilization of existing baseload capacity. As a result, the emissions impact depends on which effect dominates. Researchers also found that higher renewable penetration increases the emissions-reduction potential of data center flexibility, while lower shares favor baseload generation and may raise emissions. Key findings highlight the importance of aligning data center flexibility with renewable deployment and regional conditions.

This report presents highlights from 2024 final birth data on key demographic and medical and healthcare indicators. The number of births, the general fertility rate (the number of births per 1,000 females ages 15–44), age-specific birth rates, primary cesarean delivery rates by age of mother, and the percentage of births covered by Medicaid by age of mother are presented. Key findings include that the number of births in the United States increased 1% from 2023 to 2024, to 3,628,934 births. The general fertility rate declined 1% from 2023 to 2024 to 53.8 births per 1,000 females ages 15–44. From 2023 to 2024, birth rates declined for females in age groups 15–34, were unchanged for women ages 35–39, and rose for women ages 40–44. The primary cesarean delivery rate increased to 22.9% in 2024. The percentage of mothers for whom Medicaid was the primary source of payment for the delivery declined overall and for each maternal age group.

Although infant mortality has declined in the United States over time, previous research shows that differences persist across geographic areas and by maternal and infant characteristics. An earlier report based on 2014 vital statistics data examined differences in infant mortality by urbanization level. This report presents trends in infant mortality among rural, small and medium metropolitan, and large metropolitan counties in the United States from 2014 through 2023, and infant mortality rates by age at death, mother’s age, and maternal race and Hispanic origin for combined years 2021–2023. Key findings include that the infant mortality rate declined from 2014 to 2020 for all urbanization levels, and then had varying trends across urbanization levels from 2020 to 2023. During 2021–2023, total infant, neonatal, and post-neonatal mortality rates were higher in rural and small and medium metropolitan counties compared with large metropolitan counties. Infant mortality rates were higher in rural and small and medium metropolitan counties compared with large metropolitan counties for infants of mothers of all age groups in 2021–2023. Infant mortality rates were higher in rural and small and medium metropolitan counties compared with large metropolitan counties for infants of most maternal race and Hispanic-origin groups.

Opportunity Passport is a financial education and matched savings program designed to help young people achieve success as they move from foster care to adulthood. Launched in 2003, the program was developed to address a need expressed by young people: a lack of opportunities to learn about finances and practice financial decision-making while in foster care. The Opportunity Passport’s main components work together to provide young people with financial literacy training, banking courses, matched savings, and goal setting. Nationally, 31% of participants who were ever enrolled in Opportunity Passport purchased assets using the matched savings, though there is substantial variation at the state level. Most participants used their savings to buy a car (32%) or pay a housing expense, such as a rental deposit (18%). In addition, across many indicators of emerging adult growth and stability, Opportunity Passport participants fare better than their peers who are not enrolled. Participants who are saving and making approved purchases are more likely to report working full time and for at least six months; attending school or work-related training; having safe, stable and affordable housing; and being able to cover monthly expenses.