How to Implement AI Strategies in the Public Sector

𝟮𝟴 𝗣𝗮𝗴𝗲𝘀. 𝟯 𝗣𝗶𝗹𝗹𝗮𝗿𝘀. 𝗧𝗵𝗲 𝗠𝗼𝘀𝘁 𝗔𝗰𝘁𝗶𝗼𝗻𝗮𝗯𝗹𝗲 𝗔𝗜 𝗕𝗹𝘂𝗲𝗽𝗿𝗶𝗻𝘁 𝗬𝗲𝘁. The America’s AI Action Plan (July 2025) isn’t theory—it’s execution-focused. 𝗔𝗻𝗱 𝗶𝘁’𝘀 𝗽𝗮𝗰𝗸𝗲𝗱 𝘄𝗶𝘁𝗵 𝘄𝗵𝗮𝘁’𝘀 𝗼𝗳𝘁𝗲𝗻 𝗺𝗶𝘀𝘀𝗶𝗻𝗴 𝗶𝗻 𝗰𝗼𝗿𝗽𝗼𝗿𝗮𝘁𝗲 𝗮𝗻𝗱 𝗴𝗼𝘃𝗲𝗿𝗻𝗺𝗲𝗻𝘁 𝗔𝗜 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀: 𝘱𝘳𝘢𝘤𝘵𝘪𝘤𝘢𝘭 𝘭𝘦𝘷𝘦𝘳𝘴 𝘧𝘰𝘳 𝘢𝘥𝘰𝘱𝘵𝘪𝘰𝘯 𝘢𝘵 𝘴𝘤𝘢𝘭𝘦. 𝗜𝘁 𝗰𝗲𝗻𝘁𝗲𝗿𝘀 𝗮𝗿𝗼𝘂𝗻𝗱 𝘁𝗵𝗿𝗲𝗲 𝗽𝗶𝗹𝗹𝗮𝗿𝘀: Innovation, Infrastructure, and International Leadership. 𝗛𝗲𝗿𝗲 𝗮𝗿𝗲 𝘁𝗵𝗲 𝗱𝗶𝗿𝗲𝗰𝘁 𝗺𝗼𝘃𝗲𝘀 𝘁𝗵𝗮𝘁 𝗺𝗮𝘁𝘁𝗲𝗿: The plan 𝗺𝗮𝗻𝗱𝗮𝘁𝗲𝘀 𝗿𝗲𝗴𝘂𝗹𝗮𝘁𝗼𝗿𝘆 𝘀𝗮𝗻𝗱𝗯𝗼𝘅𝗲𝘀—especially for healthcare, agriculture, and energy—to speed up real-world AI use. It proposes 𝗔𝗜 𝗖𝗲𝗻𝘁𝗲𝗿𝘀 𝗼𝗳 𝗘𝘅𝗰𝗲𝗹𝗹𝗲𝗻𝗰𝗲 to help sectors test and share learnings across open environments. It 𝗰𝗮𝗹𝗹𝘀 𝗳𝗼𝗿 𝗼𝗽𝗲𝗻-𝘄𝗲𝗶𝗴𝗵𝘁 𝗺𝗼𝗱𝗲𝗹 𝘀𝘂𝗽𝗽𝗼𝗿𝘁—so startups, universities, and enterprises with sensitive data don’t get locked out. It 𝗱𝗲𝗺𝗮𝗻𝗱𝘀 𝗔𝗜 𝗮𝗰𝗰𝗲𝘀𝘀 𝗮𝗻𝗱 𝘁𝗿𝗮𝗶𝗻𝗶𝗻𝗴 𝗳𝗼𝗿 𝗮𝗹𝗹 𝗳𝗲𝗱𝗲𝗿𝗮𝗹 𝗲𝗺𝗽𝗹𝗼𝘆𝗲𝗲𝘀 who can benefit from it—no more excuses. It 𝗶𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝗲𝘀 𝗮 𝗳𝗲𝗱𝗲𝗿𝗮𝗹 𝗔𝗜 𝗽𝗿𝗼𝗰𝘂𝗿𝗲𝗺𝗲𝗻𝘁 𝘁𝗼𝗼𝗹𝗯𝗼𝘅 and a talent exchange for agencies to share AI skills. It 𝘁𝗮𝗿𝗴𝗲𝘁𝘀 𝗻𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝘃𝗶𝘁𝘆 𝘄𝗶𝘁𝗵 𝗱𝗼𝗺𝗮𝗶𝗻-𝘀𝗽𝗲𝗰𝗶𝗳𝗶𝗰 benchmarks—measuring AI impact on actual workflows. It 𝗽𝗿𝗶𝗼𝗿𝗶𝘁𝗶𝘇𝗲𝘀 𝘀𝗸𝗶𝗹𝗹𝗲𝗱 𝘁𝗿𝗮𝗱𝗲 𝗷𝗼𝗯𝘀—𝘄𝗶𝘁𝗵 𝘁𝗮𝘅 𝗯𝗲𝗻𝗲𝗳𝗶𝘁𝘀 𝗳𝗼𝗿 𝗔𝗜 𝘁𝗿𝗮𝗶𝗻𝗶𝗻𝗴, apprenticeships, and career-tech programs. It makes 𝗲𝗻𝗲𝗿𝗴𝘆 𝗴𝗿𝗶𝗱 𝗿𝗲𝗳𝗼𝗿𝗺 𝗮𝗻𝗱 𝗱𝗮𝘁𝗮 𝗰𝗲𝗻𝘁𝗲𝗿 𝗯𝘂𝗶𝗹𝗱𝗼𝘂𝘁 a core part of AI readiness—not just compute. It invests in 𝗶𝗻𝘁𝗲𝗿𝗽𝗿𝗲𝘁𝗮𝗯𝗶𝗹𝗶𝘁𝘆, 𝗮𝗱𝘃𝗲𝗿𝘀𝗮𝗿𝗶𝗮𝗹 𝗿𝗼𝗯𝘂𝘀𝘁𝗻𝗲𝘀𝘀, 𝗮𝗻𝗱 𝗻𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝘀𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗲𝘃𝗮𝗹𝘂𝗮𝘁𝗶𝗼𝗻𝘀 for frontier models. It 𝗰𝗮𝗹𝗹𝘀 𝗼𝘂𝘁 𝗱𝗲𝗲𝗽𝗳𝗮𝗸𝗲 𝘁𝗵𝗿𝗲𝗮𝘁𝘀 𝘁𝗼 𝗰𝗼𝘂𝗿𝘁𝘀 𝗮𝗻𝗱 𝗼𝘂𝘁𝗹𝗶𝗻𝗲𝘀 how DOJ and National Institute of Standards and Technology (NIST) should set technical and evidentiary standards. It 𝗽𝗿𝗼𝗽𝗼𝘀𝗲𝘀 𝗮 𝗳𝘂𝗹𝗹-𝘀𝘁𝗮𝗰𝗸 𝗔𝗜 𝗲𝘅𝗽𝗼𝗿𝘁 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝘆—hardware, models, values—offered to allies and partners. 𝗧𝗵𝗶𝘀 𝗽𝗹𝗮𝗻 𝗶𝘀 𝗱𝗶𝗿𝗲𝗰𝘁, 𝘂𝗻𝗮𝗽𝗼𝗹𝗼𝗴𝗲𝘁𝗶𝗰, 𝗮𝗻𝗱 𝗶𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻-𝗺𝗶𝗻𝗱𝗲𝗱. If you're working on AI adoption—public or private—this is a must read.

The Office of Management and Budget (OMB) has issued new mandates regarding government agencies' use of artificial intelligence (AI). One of the most significant mandates is for agencies to publicly report their AI usage, including the risks involved and how they manage them. This move will help ensure transparency and accountability in AI usage.   One critical area that needs to be adequately addressed is the federal government's procurement processes to execute AI mandates. Procurement holds significant sway over markets, particularly in their early stages of development. As industry leaders debate setting standards for these technologies, the government should work with industry to help establish a baseline for these crucial discussions.   AI procurement frameworks should also be established that prescribe the terms and conditions applying to any subsequent AI contract and allow the pre-vetting of providers against predefined criteria that should include ethical requirements and risk management.   Pre-market engagement is also often essential in helping the government describe the problem and narrow down the tasks that AI may be able to assist with. This approach will allow the government to work better with potential suppliers, communicate what is being asked for and why, and highlight where the gaps are. Effective RFIs are fine, but real engagement through one-on-ones and listening sessions with industry instead of just paper-based research will impact and provide the best opportunity to understand and mitigate potential AI risks.   Risk management starts during need identification. Industry would be wise to evolve their BD processes to help the government better understand risk profiles and ethical AI earlier in the process.   #AI #Government #Accountability #Transparency #Procurement #Frameworks #AIimplementation #premarketengagement #artificialintelligence #businessstrategy #govcon

The American AI Action Plan is out. And it's a lot. Clocking in at over 70 pages, it lays out the federal government’s strategy to lead the global AI race. One of the most practical sections focuses on Accelerating AI Adoption in Government. The premise is simple: AI can help government work better and faster. Think fewer manual processes, better service delivery, and less red tape. To get there, the plan recommends: - Locking in leadership: Formalize the Chief AI Officers Council as the lead body coordinating AI efforts across government. - Sharing talent: Launch a federal talent exchange to rapidly deploy AI experts (like data scientists and engineers) where they’re most needed. - Fixing procurement: GSA will build an AI procurement toolbox, giving agencies standardized, flexible access to models that meet privacy and transparency standards. - Spreading good ideas faster: Create a system to quickly share effective AI tools and use cases across agencies. - Expanding access: Make sure employees who can benefit from advanced AI tools actually have them and know how to use them. - Piloting real impact: Use High Impact Service Providers to test how AI can directly improve service delivery to the public. The full plan is here: https://lnkd.in/enKmKkBP

AI Procurement Checklists For better or worse, governments run on procurement (if you can't figure out how to buy it, you are never going to get to use it). V interesting new paper (from David Eaves and others) looking at AI procurement in government (see below). My quick summary of best practices is something like: -Mandating expert involvement: This is a complex area, we need expert involvement -Closing loopholes: Checklists should be required for all AI systems, even those developed in-house, below certain cost thresholds, or procured without AI being initially contemplated. This universal application helps close loopholes that allow some systems to bypass rigorous scrutiny. Even a "trivial system" has the capability to create reputational damage, as several ogranizations have already discovered. Prioritizing substantive and procedural transparency: This openness allows external experts and the public to identify potential issues missed by internal reviews, fostering greater accountability and trust. Though might not always be practical due to cybersecuity and/or provacy reasons. Risk allocation (liability) between government and vendors needs to be thoughtfully apportioned to incentivize responsible - general rule of the thumb, risk should be borne by whoever is best situated to manage that risk (but, easy to say hard to do) Leveraging cross-sector partnerships: Partnerships between government, industry, academia, and civil society can help expand AI expertise, align on audit standards, identify loopholes, and understand AI limitations. The government can lead in setting standards for vetting AI systems, influencing standards in the private sector. Standardizing AI audits: These can't be entirely black boxes we need some what to have audit trails to understand what is happening, what constitutes success, etc. (Note the parts of the summary above that are helpful and accurate are courtesy of the authors, anything that is inaccurate or wrong is my fault :-) ) #procurement #AI

🚀 Navigating Generative AI in the Public Sector: Tackling 'Faucets' and 'Floods' 🚀 As an AI Evangelist in the public sector, a frequent question I address is where to start with Generative AI. The answer lies in understanding and harnessing the dynamics of 'Faucets' and 'Floods' in our workflows. 🔧 Faucets: Efficiently Managing Routine Workflows 'Faucets' represent areas with steady, manageable tasks. These are ideal for introducing Generative AI, where it can effectively handle routine tasks, like document processing or public queries, ensuring that data privacy and security are not compromised. This approach allows us to gently integrate AI, building trust and familiarity within our teams. 🌊 Floods: Essential Areas for Generative AI On the other hand, 'Floods' are scenarios where the volume of work exceeds our current capacity. This could be due to emergency situations or simply because of understaffing. In these cases, Generative AI becomes almost essential. It can efficiently manage the overflow, provided we rigorously address data security and risk mitigation. The focus is on smartly employing AI to support our teams without compromising on safety and ethical standards. 📉 Addressing Public Sector Employment Challenges With the decline in public sector employment, it's increasingly vital that Generative AI supports our existing staff. It's not just about managing current workloads; it's about scalability and future-proofing our operations. Generative AI can be a powerful tool in augmenting the capabilities of our workforce, helping them do more with less, and ensuring that public services remain responsive and effective. 💬 Join the Conversation on Responsible AI Implementation What are your perspectives on introducing Generative AI in public sector workflows? How can we ensure it's done responsibly, keeping pace with changing workforce dynamics? I invite you to share your thoughts and experiences. Let’s explore together how Generative AI can be a strategic asset in our public sector missions, focusing on sustainable, ethical, and effective solutions. 💡 #GenerativeAI #PublicSectorChallenges #FutureOfWork #ResponsibleInnovation

The new GAO report on the use of AI by federal agencies covers the current and planned uses of AI, the extent of comprehensive and accurate AI reporting by agencies, and compliance with national policy and guidance on AI. This is a must-read report for anyone working in GovCon and supporting our client's missions. The report highlights the potential of AI to transform government operations and addresses challenges in its adoption. It also includes an examination of federal laws and executive actions guiding AI implementation, agency roles in AI oversight, and recommendations for improving AI governance and use. My key takeaways: - Federal agencies are increasingly implementing AI, with about 1200 current and planned use cases. - Agencies face issues developing comprehensive and accurate AI inventories, highlighting the need for better data management and reporting. - The need for a unified definition of AI and its broad range of applications across agencies indicates the versatility and complexity of AI technologies. - Agencies show varying levels of compliance with federal AI policies and guidance, suggesting the need for standardized practices. - AI holds significant promise for transforming government operations, with examples of successful implementations improving efficiency and decision-making. For senior AI and data scientist leaders, the call to action is clear: - Provide expertise to help agencies formulate comprehensive AI strategies and roadmaps. - Offer solutions for better data management and accurate reporting of AI use cases (look at the graphic; there is an impressive number of use cases). - Work with your clients aligning with federal AI policies and directives, ensuring compliance and ethical considerations, and creating actionable frameworks. - Develop training programs for agency personnel to enhance their AI-related skills and knowledge. The need to re-skill the workforce is critical. - Innovate and provide advanced AI solutions tailored to different agencies' specific needs and challenges. Create and deploy Gen AI solutions using diverse LLMs and other Gen AI techniques. #ArtificialIntelligence #GovernmentContracting #ManagementConsulting #DigitalTransformation #AIInGovernment #DataAnalytics #Innovation #TechTrends #PublicSectorAI #AIImplementation Graphic from GAO report at https://lnkd.in/g3Bf55su

AI Leaders and Product Managers! 👋👋 Exciting news on the GenAI (Generative AI) front! 🚀 Governor Gavin Newsom of California just signed an executive order with major ramifications for the future of artificial intelligence. As a significant hub for GenAI, California intends to harness the technology's potential while also safeguarding against potential risks. 🛡️💻 Governor Gavin Newsom stressed the transformative potential of GenAI and affirmed California's leadership role in shaping the future of ethical, transparent, and trustworthy AI. He emphasized a measured and responsible approach in evaluating AI's development and deployment within state government. 🏛️🧪 Key aspects of the executive order: 1. Risk-Analysis Report: The state will conduct a joint risk analysis to identify vulnerabilities and potential threats posed by GenAI to California's critical energy infrastructure. ⚠️ 2. Procurement Blueprint: Guidelines will be issued for public sector procurement, usage, and essential training for GenAI application. 📝 3. Beneficial Uses of GenAI Report: A report about the significant and beneficial uses of GenAI in the state, along with potential risks and harms, will be developed. 📊 4. Deployment and Analysis Framework: Guidelines for analyzing the impact of adopting GenAI tools on vulnerable communities will be developed. 🌐 5. State Employee Training: Agencies will provide training to state government workers regarding the use of state-approved GenAI. 💼 6. GenAI Partnership and Symposium: A partnership with the University of California, Berkeley, and Stanford University will be established to assess the impact of GenAI on California. 🤝 7. Legislative Engagement: Engagement with legislative partners and key stakeholders to develop policy recommendations for responsible use of AI. 🏦 8. Evaluate Impacts of AI on an Ongoing Basis: Periodic evaluations of the potential impact of GenAI on regulatory issues. 🎯 This measured approach to GenAI governance reinforces the need for Product Managers to be at the forefront of technology regulation, integrating ethical and transparent AI usage into their products, and staying tuned to policy enhancements. 🧭 Read the Executive Order at the Office of the Governor - California: https://lnkd.in/gYMsJBRq Read the news at: https://lnkd.in/gV5v455P #productmanagement #artificialintelligence #generativeai #responsbileai

Vice President Kamala Harris and the Office of Management and Budget (OMB) announced a new policy to ensure the federal government's safe use of AI. This policy introduces three binding requirements: ✔ Safeguards: Federal agencies must establish "concrete safeguards" to ensure AI usage prioritizes public interest and does not harm Americans' rights and safety. They need to assess, test, and monitor AI's impact, address biases, and maintain transparency. Agencies have until December 1, 2024, to implement these measures, or they must justify the continued use of AI systems. ✔ Transparency/Inventory: The second requirement calls for annual public disclosures of AI system inventories, risk assessments, and management strategies by federal agencies, except where such information could jeopardize public or governmental security. ✔ Chief AI Officers and AI Governance Boards: Each agency must appoint a Chief AI Officer within 60 days (by May 27, 2024) to oversee AI usage and establish AI Governance Boards to coordinate AI's governance across the government. "These measures aim to ensure that AI is used responsibly, securely, and in a manner that advances public interest, reflecting the White House's commitment to the ethical deployment of AI technology in government operations." #aitransparency #aitrust #accountanbility #aiofficer #ai #ethicalai #aisafeguards #trustedai Trusted AI™

Breaking Down the White House's America’s AI Action Plan The White House released America's AI Action Plan, which establishes a plan to advance U.S. leadership in artificial intelligence while securing critical systems and national security interests. Central to this strategy is the principle that: “The U.S. government has a responsibility to ensure the AI systems it relies on — particularly for national security applications — are protected against malicious inputs.” Key pillars and recommended actions include: - Building an AI Evaluations Ecosystem: Publish NIST‑led guidelines for rigorous testing and measurement of AI model reliability, convene biannual CAISI forums for best practices, and invest in AI testbeds for secure, real‑world evaluations. - Empowering the Workforce: Establish an AI Workforce Research Hub within the Department of Labor, integrate AI skill development into education and training programs, and pilot rapid retraining initiatives for workers impacted by AI adoption. - Accelerating Government Adoption: Formalize the Chief Artificial Intelligence Officer Council for interagency coordination, mandate agency access to frontier language models, and launch an Advanced Technology Transfer Program to share AI capabilities across agencies. - Strengthening Threat Intelligence: Create an AI Information Sharing and Analysis Center (AI‑ISAC) through DHS to consolidate AI‑security threat data, and incorporate AI incident response into CISA playbooks for both public and private sectors. - Promoting Secure‑by‑Design AI: Refine DOD’s Responsible AI and Generative AI frameworks, publish an IC Standard on AI Assurance under ODNI, and expand AI‑specific incident response capabilities in collaboration with NIST and federal cybersecurity agencies. For organizations across industries, this plan introduces new opportunities and finally a formal source of AI threat intelligence. Read the full plan: https://lnkd.in/eaiMscTF https://lnkd.in/er9k8RxH