Blog

MITRE’s AI Sandbox and FedScoop’s AI Report

MITRE has launched a Federal AI Sandbox to advance the development of AI tools for specific agency missions. The Sandbox will serve as a testing ground for select models, refining their performance in a secure, expert-informed environment. Among the first projects in the Sandbox is an AI tool designed to improve Medicare sustainability, showcasing the potential for AI to enhance public services.

FedScoop published a detailed analysis of 29 federal agency reports submitted to President Biden on the status of AI implementation. The report identifies the following critical challenges faced by federal agencies:

1. Poor data quality and inadequate data management programs
2. Under-skilled workforces lacking AI expertise
3. Insufficient technology for testing and evaluating AI tools

Federal Agencies Prioritize Responsible AI Adoption and Governance

At NVIDIA’s AI Summit, the Chief AI Officer of the Cybersecurity and Infrastructure Agency emphasized the importance of human oversight in AI adoption. He urged federal agencies to ensure persistent human involvement to monitor and guide AI systems responsibly.

In other federal developments, the General Services Administration has announced that it will enhance the federal digital login platform, Login.gov, with optional AI-enabled facial recognition. Agencies will be able to adopt this feature based on their specific needs.

The Department of Veterans Affairs is also taking steps to expand its AI capabilities. It recently issued a Request for Information seeking partners to help develop comprehensive AI governance frameworks, ensuring responsible integration as AI becomes more central to its operations.

New York and Tennessee Focus on AI Collaboration and Automation

In state news, Governor Kathy Hochul of New York unveiled the formation of the Empire State Consortium, a group of seven research institutions dedicated to studying the safe application of AI across science, medicine, government, and other sectors. This collaborative effort aims to leverage shared resources to advance both state and national AI goals.

In Tennessee, the state CIO shared updates on the progress of AI adoption during an interview this week. The CIO highlighted the success of a recent AI vendor day, which showcased technologies from 23 vendors. As the state evaluates which solutions to pursue, it is also expanding its use of robotic process automation (RPA). RPA has already proven to be a valuable tool, driving significant cost savings across state agencies.

Connecticut Senator Leads Effort to Coordinate State AI Regulations

Connecticut State Senator James Maroney, a leader in AI regulation, is spearheading efforts to form a national coalition of state-level lawmakers focused on AI policy. With the absence of a unified federal framework, Maroney aims to align state regulations to avoid a “patchwork” legal landscape, where differing AI laws across states could complicate compliance and innovation. His goal is for states to collaborate on passing consistent policies, ensuring a more cohesive approach to AI regulation nationwide.

NASCIO Insights: AI Use Cases and Implementation Challenges

At the National Association of State Chief Information Officers (NASCIO) conference, StateTech published brief interviews with state CIOs highlighting the opportunities and challenges of AI adoption. These conversations provide a snapshot of how state leaders are navigating AI technology, with videos covering:

• Top AI Use Cases
• Top Challenges Facing AI Implementation
• Top Priorities for the Contemporary CIO

AI in Disaster Relief: U.S. Military’s AI Strategy in Hurricane Response

StateScoop reports that AI played a pivotal role in the U.S. military’s response to Hurricane Helene. AI-driven tools were utilized to manage resources effectively and target the most impacted communities, enabling swift and strategic relief efforts.

You can read the full article here.

The U.S. Department of Health and Human Services (HHS) is developing a new AI strategic plan, scheduled for release in January 2025, to enhance the agency’s operations across the healthcare and human services sectors. Micky Tripathi, HHS’s chief AI officer, outlined the plan’s broad scope at the NVIDIA AI Summit. He emphasized its role in areas ranging from healthcare delivery to public health. The strategy aims to integrate AI while balancing innovation with regulatory oversight, fostering internal efficiencies and external support for healthcare innovators.

Tripathi noted the department’s focus on ensuring AI drives innovation and establishes necessary guardrails to prevent unintended consequences. While promoting innovation in the private sector, HHS also intends to identify areas where government should step back and allow markets to lead. The strategic plan aims to reflect the evolving needs of the healthcare sector and define the government’s role in supporting or limiting AI use, ensuring responsible innovation without stifling progress.

Why Read This Article?

Federal AI Strategy: Insight into one of the largest federal AI deployments and how it aligns with public service goals.

Ethical Guardrails: Emphasis on creating ethical “guardrails” for AI

Balance of Regulation & Innovation: Illustrates a government-led strategy to manage the interplay between innovation and regulation

This week brings several significant developments at the intersection of government and AI technology, highlighting how states and cities across the U.S. are advancing AI policy, data infrastructure, and education.

Updated AI Policy Tracker Map

Government Technology has published an updated AI policy tracker map, marking a major shift in state-level AI governance. By early October, 33 states have formed AI Task Forces or equivalent bodies, up from just 26 earlier this year. This rapid growth shows an increasing commitment to formalizing nationwide AI oversight. The map also highlights which states have implemented significant AI regulations or governance policies, which are developing AI literacy and training curriculums for both public and private sector workforces. These developments reflect the growing need for structured, responsible AI integration at all levels of government.

New York City’s Data Modernization Efforts

New York City’s Office of Technology and Innovation has announced a major initiative to consolidate, reorganize, and update its city-wide datasets. This project is designed to optimize the city’s data infrastructure, enabling more effective deployment of AI systems. By modernizing and streamlining these datasets, New York is setting the stage for AI-driven improvements in city operations and public services, offering a model for other cities looking to harness AI’s potential.

California’s AI Literacy Law for K-12 Schools

In a groundbreaking move, California has passed new legislation requiring K-12 schools to develop and implement AI literacy curriculums. The law mandates that students be trained to use AI across key subjects such as math, science, and history, ensuring that the next generation is prepared for an AI-powered future. This forward-thinking policy positions California as a leader in integrating AI into education, addressing technological competency and ethical considerations.

This past week, NASCIO (the National Association of State CIOs) held its annual conference, bringing together state CIOs to discuss some of the most pressing challenges in government IT, with AI at the forefront of the conversation. The event provided an opportunity for state leaders to exchange ideas and collectively examine how emerging technologies like AI are shaping the future of public services.

The conference culminated in the release of NASCIO’s highly anticipated annual CIO survey, which offered valuable insights into the status of AI across state governments. The survey revealed critical data on the types of AI applications in use, implementation challenges, and state agencies’ growing ambitions to integrate AI more fully into their operations.

Below are some key highlights from the conference:

• The Debate on Chief AI Officers: NASCIO’s president posed a thought-provoking question during his keynote: Is appointing Chief AI Officers at the state level more effective, or would the formation of state-wide AI oversight teams provide better governance? This conversation underlined the need for balanced leadership and thoughtful AI strategy across agencies.

• Concerns About Privacy and Legal Compliance: Many state CIOs expressed concerns over how fast AI is being adopted. They warned that AI’s deployment could lead to potential breaches of digital rights and privacy laws without robust privacy frameworks and legal safeguards.

• AI’s Impact on IT Workforces: A common thread in the discussions was the readiness of the current IT workforce to manage AI technology. CIOs highlighted the need for upskilling and increased funding to ensure that teams can handle the complexity of AI projects and maintain data integrity.

• Data Security and AI’s Role: With cybersecurity being a top priority, state CIOs acknowledged AI’s potential to enhance security. At the same time, they emphasized the importance of safeguarding government data, particularly as AI’s role expands and the threat landscape grows.

• Enterprise Architecture Maturity: Only 13% of state CIOs reported having high levels of enterprise architecture (EA) maturity in their states, sparking discussions on the necessity of well-structured policies and data management practices to support AI initiatives effectively.

State governments are already implementing AI, with notable examples from North Carolina, Colorado, and Arizona leading the charge. As these states face challenges and embrace innovations, it’s clear that the path forward for AI in the public sector will require a careful balance between innovation, security, and ethical considerations.

Why Read These Articles?

AI Strategy and Leadership: The approaches state CIOs take in AI adoption offer valuable insights into navigating similar challenges in public sector innovation.

Optimizing Data Practices: The focus on enterprise architecture and data readiness directly aligns with the need for robust data management and modernization.

Reskilling for AI: Addressing the skill gaps in IT workforces highlights the importance of training and reskilling initiatives.

Mitigating AI-Related Risks: The discussions around security and privacy concerns emphasize the critical need for risk management strategies that ensure safe and compliant AI deployments.

Federal Highlights

The U.S. Department of Labor has proactively promoted inclusive hiring practices by introducing an in-house AI framework. This new initiative focuses on reducing bias in the hiring process and improving recruitment efforts for disabled individuals. Meanwhile, over 20 federal executive branch agencies have released detailed compliance plans, following recent guidance from the Office of Budget and Management (OMB) on AI governance. These plans outline AI procurement, deployment, risk assessment, and data management strategies, signaling a more structured approach to federal AI oversight.

In addition, the General Services Administration (GSA) reported a 41% increase in enrollment for its newly launched AI safety and training curriculum for federal employees. This significant uptake reflects a growing commitment within federal agencies to equip staff with the necessary knowledge and skills to manage and deploy AI technologies safely.

State-Level News

Maryland has joined New Jersey and California in rolling out an AI training curriculum tailored for public sector employees. This initiative aims to ensure that government workers are prepared to engage with AI technologies in their daily operations effectively, enhancing overall state agency efficiency.

In New York, the city’s sixth annual Transit Tech Lab Competition concluded, bringing together dozens of AI developers eager to pitch solutions to improve NYC’s transit systems. Of the many participants, 18 developers were selected to partner with transit authorities over the next year to pilot AI technologies to improve the city’s public transportation efficiency.

In Other News

A bipartisan group of Stanford professors released a series of essays titled “The Digitalist Papers,” offering diverse viewpoints on AI regulation in the U.S. Among the contributors, former Secretary of State Condoleezza Rice advocated for more public involvement in shaping the role of AI in government. In contrast, others argued for minimal government regulation of AI development.

Lastly, Prepared, a leading supplier of emergency dispatch technology, raised $27 million in funding to develop AI tools that enhance 911’s rapid response capabilities. Meanwhile, StateTech published an insightful op-ed this week, exploring how federal agencies can leverage AI to design more accessible programs and spaces, focusing on improving accessibility for disabled individuals.

Governor Gavin Newsom of California vetoed a landmark AI regulation bill to establish comprehensive safety measures for artificial intelligence (AI). The bill, SB 1047, intended to establish regulations on large-scale AI systems, faced criticism for its broad scope, with concerns about the uniform standards across both high- and low-risk AI applications.

In his veto statement, Newsom emphasized that the bill failed to differentiate between AI use cases involving sensitive data and critical decision-making versus more routine tasks, and that a one-size-fits-all approach could stifle innovation. While the broader public discourse will evolve, we present four perspectives from leading media outlets—The Associated Press, The New York Times, The Wall Street Journal, and The Financial Times—which offer unique insights into the implications of this decision.

Why Read These Articles?

Regulatory Implications for AI in Public Sector: The debate around SB 1047 can influence how future AI governance is shaped, particularly for government agencies.

Balancing Innovation with Risk: These articles delve into the challenge of promoting AI innovation while ensuring public safety, a key concern for anyone involved in AI governance.

Understanding Risk-Based AI Governance: The discussion on SB 1047’s failure to differentiate between high- and low-risk AI applications highlights the importance of tailoring AI regulations to specific use cases.

Shaping Future AI Policy: California often sets the stage for broader tech policy trends.

State and Local AI Adoption

EY recently released a nationwide survey revealing that state and local agencies are lagging behind their federal counterparts in adopting AI technologies. The survey found that only 51% of state and local agencies use AI daily, compared to 64% of federal agencies. It also highlighted the major challenges to wider AI adoption, including a lack of governance and ethical frameworks, inadequate data infrastructure, and uncertainty about which AI tools best meet local needs.

These obstacles took center stage at the Digital Benefits Conference hosted by the Beeck Center for Social Impact and Innovation. Leaders from local and state governments discussed how the absence of proper governance and data quality programs is holding back AI advancements in public benefit programs.

Indiana’s AI Innovations

Indiana stood out this week with two major AI-related announcements. The state has become one of the first to implement a generative AI chatbot on its official website, designed to guide users through state services and programs. Officials emphasized the importance of rigorous safety testing, mindful of lessons learned from previous AI missteps in other regions, like New York City.

Additionally, Indiana announced its plans to integrate an AI assistant into the “Pivot” software, a tool that connects job seekers with resources across six state agencies. The AI will help tailor recommendations for job seekers, though concerns about the need for a robust data quality program remain.

Federal AI Updates

In Washington, the Committee on House Administration and the Office of the Chief Administrative Officer introduced a House-wide policy for safely deploying AI. This framework includes ethical guidelines and approved use cases, ensuring responsible AI use by House members and offices.

Meanwhile, a new bill progressing through the Senate could require federal agencies to designate officials to oversee service delivery improvements, including leveraging AI to enhance government efficiency.

However, not all federal AI developments are being met with optimism. The U.S. Commission on Civil Rights chair raised concerns about the risks associated with AI-driven facial recognition technologies, calling for federal oversight to prevent civil rights violations. Senators Chuck Schumer and Ed Markey echoed these concerns in a letter to the White House, urging the establishment of civil rights offices in federal agencies using AI for significant decision-making.

AI Safety on the Global Stage

President Biden announced plans for an international AI Safety Summit in San Francisco post-election, where government leaders and experts from the U.S. and the EU will gather to discuss unified AI safety policies. This signals an increasing push toward global cooperation on AI governance.

AI Legislation

Government Technology took a deep dive into the 120+ AI-related bills currently making their way through Congress. Most of these focus on AI’s applications in science, commerce, and defense, with few addressing the need for governance or regulatory frameworks to curb algorithmic bias in the public and private sectors.

In California, debate continues over a broad AI regulatory bill. An op-ed published by Government Technology urged Governor Newsom to embrace regulation now, even if imperfect, and refine the law as the state gains more practical experience with AI technologies.

Building Trust Through AI in Colorado’s New Data Governance Framework

Colorado’s newly released “Guide to Artificial Intelligence” aims to support state agencies in implementing generative AI technologies. Amy Bhikha, the state’s Chief Data Officer, emphasizes that success will depend on fostering innovation, building trust, and enhancing the state’s data governance program. The guidelines focus on balancing innovation with trust through a robust data governance framework, which involves inventorying state data and ensuring its quality and bias-free nature. This initiative follows earlier AI legislation, SB 24-205, establishing consumer protections for AI systems. Bhikha also highlighted the need for AI literacy among state employees and collaboration with other states to shape AI policies.

NASCIO Survey Reveals Gaps in AI Data Management Across States

A recent National Association of State CIOs (NASCIO) survey reveals that while 95% of 46 state respondents recognize the growing significance of AI and generative AI in data management, only 22% have implemented dedicated data quality programs. The survey focuses on data governance and business intelligence, but areas such as metadata and master data management are under-prioritized. NASCIO’s executive director, Doug Robinson, stresses the need for robust data quality frameworks and a data-centric culture to drive innovation and improve citizen services. The report also identifies a skills gap, with many agencies lacking data stewards and data literacy managers.

New Legislation Aims to Increase Transparency in New York City’s AI Use

New York is taking significant steps to enhance the transparency and responsible use of artificial intelligence (AI) in city and state governments. In New York City, Councilmember Jennifer Gutiérrez is introducing new legislation requiring publishing a public list of approved AI tools used by city agencies. This effort aims to boost transparency and accountability by providing detailed information on how AI tools are used and how they handle data. Scheduled for release in February 2025, the list will be updated every six months to foster collaboration between city agencies and ensure responsible use of AI technologies.

At the state level, the New York State Forum has established an AI Workgroup led by Gail Galusha, Director of Data and AI Governance. This workgroup, which includes both public and private sector members, is tasked with deepening the understanding of AI in the public sector, promoting ethical AI practices, and equipping state governments with tools for AI implementation. Both initiatives reflect New York’s leadership in the evolving landscape of AI governance and responsible technology use.

New AI Infrastructure Task Force to Boost US Leadership and Clean Energy Initiatives

The White House announced the creation of a new task force on AI data center infrastructure following a meeting with top executives from major AI and tech companies, including OpenAI, Nvidia, and Anthropic. The task force will focus on securing clean energy, workforce needs, and accelerating the construction of AI data centers to strengthen U.S. leadership in AI. The initiative aims to support national security, create jobs, and ensure the infrastructure is powered by renewable energy. However, civil society groups expressed concerns over environmental impact and transparency, warning of potential negative consequences for energy costs and climate goals.

House Panel Passes AI Bills to Boost Research, Workforce, and Safety Initiatives

The House Science, Space, and Technology Committee passed nine bipartisan AI-related bills, including legislation to codify the National AI Research Resource (NAIRR) and establish the AI Safety Institute under a new name. These bills aim to enhance AI research, workforce development, and safety measures. While many House bills have counterparts moving through the Senate, differences remain, particularly around funding for AI safety initiatives. Legislators emphasize the importance of advancing AI technology without imposing burdensome regulations, but concerns over underfunding and international competition remain key discussion points.

You can read the full article here.

Salesforce has introduced an “Industries AI” initiative designed to cater to specific sectors, including government. This project provides AI tools to improve public sector operations, particularly in areas such as benefit application processing, digital assessments, and inspection preparation. These tools aim to streamline processes, reduce administrative burden, and enhance decision-making. The initiative covers other industries like energy, education, and healthcare, signals Salesforce’s growing focus on public sector needs, and highlights AI’s role in transforming government operations.

Why Read This Article?

Targeted AI tools for government: The article provides insights into AI solutions specifically designed to address public sector needs.

Real-world use cases: The article highlights actionable examples, such as benefits processing and inspection tools.

Innovation in government technology: Understanding new AI technologies that large companies like Salesforce are investing in can help leaders anticipate future trends.

State of California Seeks AI Solutions for Public Sector Problems

California is turning to generative AI (GenAI) to tackle some of its most pressing challenges, including housing, individuals without housing, and budget analysis. To explore how large language models (LLMs) might assist, the state is hosting a GenAI showcase, where vendors will demonstrate cutting-edge capabilities. The initiative follows Governor Gavin Newsom’s executive order and summit to examine GenAI’s potential to enhance public services. Vendors will have the opportunity to present AI solutions to key state departments to address inefficiencies and improve service delivery for residents. The first showcase will focus on California’s housing crisis, inviting AI developers to present their technology to the state government for use case evaluation and possible acquisition. 

Presidential Advisors Advocate for AI Testing Protocols in Policing

An advisory panel to the president has approved recommendations requiring federal law enforcement agencies to follow standardized protocols when field-testing AI tools. The National AI Advisory Committee’s Law Enforcement Subcommittee proposed a checklist for testing AI, emphasizing transparency and performance measurement. The recommendations aim to establish clear guidelines for AI testing, ensuring tools are safe and effective before broader adoption. The subcommittee also urged publicizing testing results and advocated for additional funding to support state and local law enforcement’s AI testing efforts. These steps mark progress toward responsible AI integration in law enforcement.

Maryland’s Intentional Approach to AI Technology

Maryland officials are cautiously approaching AI due to concerns over data security and unpredictable software behavior. While acknowledging AI’s potential to streamline services, officials stress the importance of safeguarding sensitive information. Governor Wes Moore’s executive order supports AI’s benefits but calls for strict oversight, and state agencies are closely monitoring AI’s use to prevent unauthorized data sharing. Local leaders also ensure AI applications are carefully tested before broader implementation to protect public trust in government operations.

Civil Rights Groups Challenge DHS’s Use of AI in Immigration Enforcement

A coalition of over 140 immigrant and civil rights organizations has sent a letter to Secretary of Homeland Security Alejandro Mayorkas, raising concerns about the Department of Homeland Security’s (DHS) use of artificial intelligence. The letter calls for DHS to suspend certain AI systems, particularly those used by Customs and Border Protection and Immigration and Customs Enforcement, citing violations of federal policies on responsible AI use. As agencies face upcoming deadlines related to AI use case inventories, the coalition argues that AI tools deployed for immigration enforcement lack transparency and may perpetuate bias and discrimination.

A Comprehensive Guide to Data Policy for AI Development

The Data Foundation has released a guide aimed at helping policymakers navigate the challenges of data policy in the context of artificial intelligence. As AI systems heavily rely on vast amounts of data, the guide emphasizes the importance of high-quality, responsibly governed data. It highlights key components for sound data practices, including data integrity, privacy protections, transparency, and technical infrastructure. The guide also addresses the evolving landscape of AI, urging policymakers to develop comprehensive approaches to ensure AI data use aligns with public interests, ethical standards, and democratic values.