[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-how-state-lawmakers-are-using-ai-to-research-fact-check-and-draft-legislation-en":3,"ArticleBody_dsX4qbnXmGYa9QvXgZI1SNLP4kKm0E1kbBcrp53cdZg":216},{"article":4,"relatedArticles":187,"locale":66},{"id":5,"title":6,"slug":7,"content":8,"htmlContent":9,"excerpt":10,"category":11,"tags":12,"metaDescription":10,"wordCount":13,"readingTime":14,"publishedAt":15,"sources":16,"sourceCoverage":58,"transparency":60,"seo":63,"language":66,"featuredImage":67,"featuredImageCredit":68,"isFreeGeneration":72,"trendSlug":73,"niche":74,"geoTakeaways":78,"geoFaq":87,"entities":97},"69f259ada569d797da77af45","How State Lawmakers Are Using AI to Research, Fact-Check, and Draft Legislation","how-state-lawmakers-are-using-ai-to-research-fact-check-and-draft-legislation","Statehouses must process more information with fewer people. In South Dakota, 70 part‑time legislators share roughly 60 staffers, the thinnest legislative staff in the country. [2] In that context, AI that can summarize documents, search regulations, and draft language looks like basic infrastructure, not a novelty.  \n\nModern bills are long, technical, and often shaped by language supplied by interest groups. [3] Generative models now function as “synthetic staffers,” assembling intricate statutory text in minutes instead of days.  \n\nThis is no longer hypothetical:  \n\n- In 2023, Representative [Ted Lieu](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FTed_Lieu) used [ChatGPT](\u002Fentities\u002F6939891c312dc892c4c183ff-chatgpt) to write a federal resolution on AI. [1]  \n- Federal agencies, including the [U.S. Department of Education](\u002Fentities\u002F6939892b312dc892c4c18411-u-s-department-of-education), have tested AI in regulatory drafting. [1]  \n- The [Trump administration](\u002Fentities\u002F69398a91312dc892c4c1844b-trump-administration) has reportedly eyed tools such as [Gemini](\u002Fentities\u002F693adb3d312dc892c4c187e4-gemini) for transportation rules. [1]  \n\n📊 **Key takeaway:** AI is already embedded in U.S. lawmaking; the question is how to channel it responsibly, not whether it should exist.  \n\nYounger, wealthier, more Democratic-leaning states introduce the most AI-related bills, suggesting that digital capacity and political appetite move together. [6] This article surveys why legislatures turn to AI, how they use it, and what guardrails are needed.  \n\n---\n\n## Why State Legislatures Are Turning to AI  \n\nPart-time legislators juggle multiple roles with limited staff. [Kent Roe](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FKent_Roe) in South Dakota, for example:  \n\n- Works as a farmland appraiser  \n- Serves on a utility board and church council  \n- Spends up to 40 days a year legislating, with minimal research support [2]  \n\nFor officials in this position, tools that can instantly summarize a 200‑page report or draft an amendment determine whether they can keep up.  \n\nKey pressures driving AI uptake:  \n\n- **Complexity of statutes:** Modern laws can span hundreds of pages and reflect specialized lobbying input. [3]  \n- **Lobbyist dominance:** Lobbyists often supply pre-written fragments, exemptions, and carve-outs because lawmakers lack time to draft them. [3]  \n- **Capacity gap:** AI promises to restore some drafting power to legislators and nonpartisan staff, potentially broadening who can participate in drafting.  \n\n💼 **Key point:** When internal capacity is thin, external influence—often lobbyists—fills the gap. AI offers another source of drafting labor, but not a neutral one by default.  \n\nNormalization is accelerating: after Lieu’s AI resolution, agencies and states (including [Virginia](\u002Fentities\u002F694e2fe119d266277e1497de-virginia)) began testing AI for review and drafting. [1] States most active in AI legislation—young, rich, Democratic-leaning—are also best positioned to adopt AI internally, creating a feedback loop. [6]  \n\n---\n\n## How Lawmakers Use AI for Research, Fact-Checking, and Drafting  \n\nGenerative tools are being piloted in:  \n\n- The U.S. House and Senate  \n- Foreign legislatures  \n- Local governments, including a Brazilian municipality that passed the first known AI-written law in 2023 [3]  \n\nCommon uses include:  \n\n- Searching databases and summarizing hearings  \n- Analyzing policy options  \n- Drafting bill sections and amendments [3]  \n\nStaff increasingly use AI for targeted recall rather than free-form drafting. One Midwestern staffer, for example, fed prior committee transcripts into an assistant to see how “short-term rental” had been defined in past debates, saving hours. [3]  \n\nSpecialized “regulatory operating systems” are emerging:  \n\n- **[Vulcan Technologies](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FVulcan_Centaur)** aggregates statutes, regulations, and court decisions across governments. It can analyze legal language, answer queries, generate draft guidance, and propose text with citations. [1]  \n- **Virginia** has mandated Vulcan’s use across agencies to review and streamline rules, aiming to cut one-third of regulations. [1]  \n\n💡 **Key takeaway:** The frontier is shifting from generic chatbots to domain-specific legal copilots fluent in statutes, agencies, and case law.  \n\nGiven the scale of regulation—the [Code of Federal Regulations](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FCode_of_Federal_Regulations) exceeds 200,000 pages across 200 volumes [4]—AI, with careful prompting and oversight, can:  \n\n- Identify regulations linked to particular statutes or sections  \n- Map relationships between rules, agencies, and enabling laws  \n- Answer questions like “How many regulations reference 44 U.S.C. §§ 3501–3521?” [4]  \n\nFor investigations and oversight, generative systems can:  \n\n- Synthesize large evidence sets and cluster facts  \n- Flag inconsistencies and possible misinformation  \n- Support due diligence and monitor legal or regulatory changes over time [7]  \n\nThey can draft timelines and highlight gaps, but outputs must be checked against the record. [7]  \n\nProfessional platforms like [CoCounsel Legal](\u002Fentities\u002F69e75ad86db79d4361e212bf-cocounsel-legal) now integrate research, document analysis, and drafting into one workflow, reporting about 98% accuracy on supported tasks. [8] Adoption of generative tools among legal professionals rose from 14% in 2024 to 26% in 2025, signaling similar integrated tools for legislatures. [8]  \n\n---\n\n## Risks, Oversight, and Best Practices for AI-Assisted Lawmaking  \n\nAI-written law raises constitutional as well as technical issues. Cheap, detailed drafting lets legislators:  \n\n- Write more prescriptive statutes  \n- Narrow the discretion of executive agencies that traditionally flesh out vague laws through rulemaking [3]  \n\n⚠️ **Key point:** Faster drafting can quietly reshape separation of powers by hard-coding more policy detail into statutes. [3]  \n\nQuality risks include hallucinations, omitted caveats, and misstated precedent. [7] Responsible use requires workflows where:  \n\n- Outputs are treated as hypotheses, not facts  \n- Every substantive claim is checked against primary sources  \n- Citations and cross-references are systematically verified [7]  \n\nVerification must rely on **lateral reading**:  \n\n- Ask “who can confirm this?” not “who wrote this?” [5]  \n- Break responses into discrete claims  \n- Check each claim against trusted legal databases, government publications, and reputable analyses [5]  \n\nStatehouses should adopt policies that:  \n\n- Require disclosure when AI substantially contributes to bill text  \n- Set minimum human review standards before introduction  \n- Vet tools for ethics, security, and data governance  \n- Provide training that pairs technical skills with critical evaluation and lateral reading practices [5][8]  \n\nStates already [leading on AI bills](\u002Farticle\u002Fhow-the-white-house-ai-policy-could-override-state-laws-and-reshape-global-tech-governance) are well-positioned to pilot governance frameworks for AI-assisted lawmaking, using early policy experience to guide transparent, accountable internal use. [6]  \n\n---\n\n## Conclusion: Treat AI as an Assistant, Not a Ghostwriter  \n\nAI is rapidly embedding itself in state-level lawmaking—from navigating the 200,000‑page CFR to analyzing evidence and drafting statutory language. [1][4][7] The benefits are speed and depth; the dangers are opacity, overreach, and subtle error.  \n\nLegislative leaders should map where AI is already used, set review and disclosure rules, and train staff in verification disciplines like lateral reading. [5] AI should function as a visible, accountable assistant whose work is always checked, attributed, and subject to democratic scrutiny.","\u003Cp>Statehouses must process more information with fewer people. In South Dakota, 70 part‑time legislators share roughly 60 staffers, the thinnest legislative staff in the country. \u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa> In that context, AI that can summarize documents, search regulations, and draft language looks like basic infrastructure, not a novelty.\u003C\u002Fp>\n\u003Cp>Modern bills are long, technical, and often shaped by language supplied by interest groups. \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa> Generative models now function as “synthetic staffers,” assembling intricate statutory text in minutes instead of days.\u003C\u002Fp>\n\u003Cp>This is no longer hypothetical:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>In 2023, Representative \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FTed_Lieu\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">Ted Lieu\u003C\u002Fa> used \u003Ca href=\"\u002Fentities\u002F6939891c312dc892c4c183ff-chatgpt\">ChatGPT\u003C\u002Fa> to write a federal resolution on AI. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>Federal agencies, including the \u003Ca href=\"\u002Fentities\u002F6939892b312dc892c4c18411-u-s-department-of-education\">U.S. Department of Education\u003C\u002Fa>, have tested AI in regulatory drafting. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>The \u003Ca href=\"\u002Fentities\u002F69398a91312dc892c4c1844b-trump-administration\">Trump administration\u003C\u002Fa> has reportedly eyed tools such as \u003Ca href=\"\u002Fentities\u002F693adb3d312dc892c4c187e4-gemini\">Gemini\u003C\u002Fa> for transportation rules. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>📊 \u003Cstrong>Key takeaway:\u003C\u002Fstrong> AI is already embedded in U.S. lawmaking; the question is how to channel it responsibly, not whether it should exist.\u003C\u002Fp>\n\u003Cp>Younger, wealthier, more Democratic-leaning states introduce the most AI-related bills, suggesting that digital capacity and political appetite move together. \u003Ca href=\"#source-6\" class=\"citation-link\" title=\"View source [6]\">[6]\u003C\u002Fa> This article surveys why legislatures turn to AI, how they use it, and what guardrails are needed.\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>Why State Legislatures Are Turning to AI\u003C\u002Fh2>\n\u003Cp>Part-time legislators juggle multiple roles with limited staff. \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FKent_Roe\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">Kent Roe\u003C\u002Fa> in South Dakota, for example:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Works as a farmland appraiser\u003C\u002Fli>\n\u003Cli>Serves on a utility board and church council\u003C\u002Fli>\n\u003Cli>Spends up to 40 days a year legislating, with minimal research support \u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>For officials in this position, tools that can instantly summarize a 200‑page report or draft an amendment determine whether they can keep up.\u003C\u002Fp>\n\u003Cp>Key pressures driving AI uptake:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Complexity of statutes:\u003C\u002Fstrong> Modern laws can span hundreds of pages and reflect specialized lobbying input. \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Lobbyist dominance:\u003C\u002Fstrong> Lobbyists often supply pre-written fragments, exemptions, and carve-outs because lawmakers lack time to draft them. \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Capacity gap:\u003C\u002Fstrong> AI promises to restore some drafting power to legislators and nonpartisan staff, potentially broadening who can participate in drafting.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>💼 \u003Cstrong>Key point:\u003C\u002Fstrong> When internal capacity is thin, external influence—often lobbyists—fills the gap. AI offers another source of drafting labor, but not a neutral one by default.\u003C\u002Fp>\n\u003Cp>Normalization is accelerating: after Lieu’s AI resolution, agencies and states (including \u003Ca href=\"\u002Fentities\u002F694e2fe119d266277e1497de-virginia\">Virginia\u003C\u002Fa>) began testing AI for review and drafting. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa> States most active in AI legislation—young, rich, Democratic-leaning—are also best positioned to adopt AI internally, creating a feedback loop. \u003Ca href=\"#source-6\" class=\"citation-link\" title=\"View source [6]\">[6]\u003C\u002Fa>\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>How Lawmakers Use AI for Research, Fact-Checking, and Drafting\u003C\u002Fh2>\n\u003Cp>Generative tools are being piloted in:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>The U.S. House and Senate\u003C\u002Fli>\n\u003Cli>Foreign legislatures\u003C\u002Fli>\n\u003Cli>Local governments, including a Brazilian municipality that passed the first known AI-written law in 2023 \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Common uses include:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Searching databases and summarizing hearings\u003C\u002Fli>\n\u003Cli>Analyzing policy options\u003C\u002Fli>\n\u003Cli>Drafting bill sections and amendments \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Staff increasingly use AI for targeted recall rather than free-form drafting. One Midwestern staffer, for example, fed prior committee transcripts into an assistant to see how “short-term rental” had been defined in past debates, saving hours. \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>Specialized “regulatory operating systems” are emerging:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>\u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FVulcan_Centaur\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">Vulcan Technologies\u003C\u002Fa>\u003C\u002Fstrong> aggregates statutes, regulations, and court decisions across governments. It can analyze legal language, answer queries, generate draft guidance, and propose text with citations. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Virginia\u003C\u002Fstrong> has mandated Vulcan’s use across agencies to review and streamline rules, aiming to cut one-third of regulations. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>💡 \u003Cstrong>Key takeaway:\u003C\u002Fstrong> The frontier is shifting from generic chatbots to domain-specific legal copilots fluent in statutes, agencies, and case law.\u003C\u002Fp>\n\u003Cp>Given the scale of regulation—the \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FCode_of_Federal_Regulations\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">Code of Federal Regulations\u003C\u002Fa> exceeds 200,000 pages across 200 volumes \u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>—AI, with careful prompting and oversight, can:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Identify regulations linked to particular statutes or sections\u003C\u002Fli>\n\u003Cli>Map relationships between rules, agencies, and enabling laws\u003C\u002Fli>\n\u003Cli>Answer questions like “How many regulations reference 44 U.S.C. §§ 3501–3521?” \u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>For investigations and oversight, generative systems can:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Synthesize large evidence sets and cluster facts\u003C\u002Fli>\n\u003Cli>Flag inconsistencies and possible misinformation\u003C\u002Fli>\n\u003Cli>Support due diligence and monitor legal or regulatory changes over time \u003Ca href=\"#source-7\" class=\"citation-link\" title=\"View source [7]\">[7]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>They can draft timelines and highlight gaps, but outputs must be checked against the record. \u003Ca href=\"#source-7\" class=\"citation-link\" title=\"View source [7]\">[7]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>Professional platforms like \u003Ca href=\"\u002Fentities\u002F69e75ad86db79d4361e212bf-cocounsel-legal\">CoCounsel Legal\u003C\u002Fa> now integrate research, document analysis, and drafting into one workflow, reporting about 98% accuracy on supported tasks. \u003Ca href=\"#source-8\" class=\"citation-link\" title=\"View source [8]\">[8]\u003C\u002Fa> Adoption of generative tools among legal professionals rose from 14% in 2024 to 26% in 2025, signaling similar integrated tools for legislatures. \u003Ca href=\"#source-8\" class=\"citation-link\" title=\"View source [8]\">[8]\u003C\u002Fa>\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>Risks, Oversight, and Best Practices for AI-Assisted Lawmaking\u003C\u002Fh2>\n\u003Cp>AI-written law raises constitutional as well as technical issues. Cheap, detailed drafting lets legislators:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Write more prescriptive statutes\u003C\u002Fli>\n\u003Cli>Narrow the discretion of executive agencies that traditionally flesh out vague laws through rulemaking \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>⚠️ \u003Cstrong>Key point:\u003C\u002Fstrong> Faster drafting can quietly reshape separation of powers by hard-coding more policy detail into statutes. \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>Quality risks include hallucinations, omitted caveats, and misstated precedent. \u003Ca href=\"#source-7\" class=\"citation-link\" title=\"View source [7]\">[7]\u003C\u002Fa> Responsible use requires workflows where:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Outputs are treated as hypotheses, not facts\u003C\u002Fli>\n\u003Cli>Every substantive claim is checked against primary sources\u003C\u002Fli>\n\u003Cli>Citations and cross-references are systematically verified \u003Ca href=\"#source-7\" class=\"citation-link\" title=\"View source [7]\">[7]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Verification must rely on \u003Cstrong>lateral reading\u003C\u002Fstrong>:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Ask “who can confirm this?” not “who wrote this?” \u003Ca href=\"#source-5\" class=\"citation-link\" title=\"View source [5]\">[5]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>Break responses into discrete claims\u003C\u002Fli>\n\u003Cli>Check each claim against trusted legal databases, government publications, and reputable analyses \u003Ca href=\"#source-5\" class=\"citation-link\" title=\"View source [5]\">[5]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Statehouses should adopt policies that:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Require disclosure when AI substantially contributes to bill text\u003C\u002Fli>\n\u003Cli>Set minimum human review standards before introduction\u003C\u002Fli>\n\u003Cli>Vet tools for ethics, security, and data governance\u003C\u002Fli>\n\u003Cli>Provide training that pairs technical skills with critical evaluation and lateral reading practices \u003Ca href=\"#source-5\" class=\"citation-link\" title=\"View source [5]\">[5]\u003C\u002Fa>\u003Ca href=\"#source-8\" class=\"citation-link\" title=\"View source [8]\">[8]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>States already \u003Ca href=\"\u002Farticle\u002Fhow-the-white-house-ai-policy-could-override-state-laws-and-reshape-global-tech-governance\" class=\"internal-link\">leading on AI bills\u003C\u002Fa> are well-positioned to pilot governance frameworks for AI-assisted lawmaking, using early policy experience to guide transparent, accountable internal use. \u003Ca href=\"#source-6\" class=\"citation-link\" title=\"View source [6]\">[6]\u003C\u002Fa>\u003C\u002Fp>\n\u003Chr>\n\u003Ch2>Conclusion: Treat AI as an Assistant, Not a Ghostwriter\u003C\u002Fh2>\n\u003Cp>AI is rapidly embedding itself in state-level lawmaking—from navigating the 200,000‑page CFR to analyzing evidence and drafting statutory language. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003Ca href=\"#source-7\" class=\"citation-link\" title=\"View source [7]\">[7]\u003C\u002Fa> The benefits are speed and depth; the dangers are opacity, overreach, and subtle error.\u003C\u002Fp>\n\u003Cp>Legislative leaders should map where AI is already used, set review and disclosure rules, and train staff in verification disciplines like lateral reading. \u003Ca href=\"#source-5\" class=\"citation-link\" title=\"View source [5]\">[5]\u003C\u002Fa> AI should function as a visible, accountable assistant whose work is always checked, attributed, and subject to democratic scrutiny.\u003C\u002Fp>\n","Statehouses must process more information with fewer people. In South Dakota, 70 part‑time legislators share roughly 60 staffers, the thinnest legislative staff in the country. [2] In that context, AI...","trend-radar",[],1020,5,"2026-04-29T19:30:48.260Z",[17,22,26,30,34,38,42,46,50,54],{"title":18,"url":19,"summary":20,"type":21},"Lawmakers are using AI to write laws. What could go wrong?","https:\u002F\u002Fwww.transformernews.ai\u002Fp\u002Fai-lawmakers-laws-vulcan-technologies-fiscalnote-policynote-virginia-vermont","Lawmakers and companies are quietly using AI to draft legislation. Experts warn the risks are underappreciated\n\nApr 09, 2026\n\nIn 2023, California congressman Ted Lieu introduced what he called the fir...","kb",{"title":23,"url":24,"summary":25,"type":21},"Artificial intelligence is creeping into American lawmaking","https:\u002F\u002Fwww.economist.com\u002Funited-states\u002F2026\u002F04\u002F23\u002Fartificial-intelligence-is-creeping-into-american-lawmaking","Apr 23rd 2026 | WASHINGTON, DC | 3 min read\n\nK ENT ROE is a busy man. In addition to his full-time job as a farmland appraiser, he sits on the board of a utility company and on the council of his loca...",{"title":27,"url":28,"summary":29,"type":21},"AI Will Write Complex Laws | Lawfare","https:\u002F\u002Fwww.lawfaremedia.org\u002Farticle\u002Fai-will-write-complex-laws","Nathan Sanders; Bruce Schneier\n\nArtificial intelligence (AI) is writing law today. This has required no changes in legislative procedure or the rules of legislative bodies—all it takes is one legislat...",{"title":31,"url":32,"summary":33,"type":21},"Unlocking the Code: How Legislators can use AI to Demystify Regulation","https:\u002F\u002Fwww.popvox.org\u002Feffective-government-fellow-projects\u002Fai-cfr","Unlocking the Code: How Legislators can use AI to Demystify Regulation\n\nNov 5\n\nWritten By POPVOX Foundation\nBY KATHRYN LUEDKE\n\nEmerging technologies are presenting promising new solutions to the seemi...",{"title":35,"url":36,"summary":37,"type":21},"Using AI Tools in Research: Fact-checking AI with Lateral Reading","https:\u002F\u002Fguides.library.tamucc.edu\u002FAI\u002FlateralreadingAI","Lateral reading: your #1 analysis tool\n\nIf you cannot take AI-cited sources at face value and you (or the AI's programmers) cannot determine where the information is sourced from, how are you going to...",{"title":39,"url":40,"summary":41,"type":21},"Analyzing the passage of state-level AI bills","https:\u002F\u002Fwww.brookings.edu\u002Farticles\u002Fanalyzing-the-passage-of-state-level-ai-bills\u002F","Artificial intelligence (AI) continues to dominate headlines, spanning chip supremacy and job losses to AI actresses and U.S. national security. These articles demonstrate that AI is top of mind acros...",{"title":43,"url":44,"summary":45,"type":21},"How to Use Generative AI for Fact Analysis and Investigation in 2025","https:\u002F\u002Fcsdisco.com\u002Fblog\u002Fgenerative-ai-for-fact-analysis-investigation","Industry & Legal Education\n\nEstimated Read Time: 7 minutes\n\nBy: DISCO\n\nPosted: May 30, 2025\n\nTable of Contents\n\nWhy use generative AI for fact analysis and investigation?\nKey features of generative AI...",{"title":47,"url":48,"summary":49,"type":21},"Legal AI tools with Westlaw and Practical Law, all in one","https:\u002F\u002Flegal.thomsonreuters.com\u002Fblog\u002Flegal-ai-tools-essential-for-attorneys\u002F","CoCounsel Legal: an all-in-one AI solution that combines GenAI and agentic AI for research, document analysis, and drafting into a single workflow.\n\nHow AI connects all the dots from legal research an...",{"title":51,"url":52,"summary":53,"type":21},"DeepSeek vs. ChatGPT: A Complete Breakdown for Small Businesses","https:\u002F\u002Foakinteractive.com\u002Fdeepseek-vs-chatgpt-a-complete-breakdown-for-small-businesses\u002F","### DeepSeek: The AI Research Powerhouse\nDeepSeek is an AI-powered research tool that excels at diving deep into the vast ocean of online information. It can rapidly analyze documents, websites, artic...",{"title":55,"url":56,"summary":57,"type":21},"The New AI Workflow: Combining ChatGPT and DeepSeek for Maximum Results","https:\u002F\u002Fdmwebsoft.com\u002Fthe-new-ai-workflow-combining-chatgpt-and-deepseek-for-maximum-results","### Introduction\n\nArtificial intelligence is revolutionizing the way companies function, offering powerful tools for simplifying processes, improving efficiency, and automating repetitive processes. T...",{"totalSources":59},10,{"generationDuration":61,"kbQueriesCount":59,"confidenceScore":62,"sourcesCount":59},326384,100,{"metaTitle":64,"metaDescription":65},"State lawmakers adopt AI tools for faster legislation","Understaffed statehouses use AI to summarize, fact-check, and draft bills. Read how legislatures use AI, the risks, and simple guardrails to ensure accuracy.","en","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1576082176859-e557bdc7b1b4?ixid=M3w4OTczNDl8MHwxfHNlYXJjaHwxfHxzdGF0ZSUyMGxhd21ha2VycyUyMHVzaW5nJTIwcmVzZWFyY2h8ZW58MXwwfHx8MTc3NzQ5MDM0OHww&ixlib=rb-4.1.0&w=1200&h=630&fit=crop&crop=entropy&auto=format,compress&q=60",{"photographerName":69,"photographerUrl":70,"unsplashUrl":71},"Katherine McAdoo","https:\u002F\u002Funsplash.com\u002F@ohaikatherine?utm_source=coreprose&utm_medium=referral","https:\u002F\u002Funsplash.com\u002Fphotos\u002Fthe-pennysylvania-state-capitol-signage-HLKNH1-ITr0?utm_source=coreprose&utm_medium=referral",true,"state-lawmakers-using-ai-to-research-fact-check-draft-legislation",{"key":75,"name":76,"nameEn":77},"ia","Intelligence Artificielle","Artificial Intelligence",[79,81,83,85],{"text":80},"AI is already embedded in U.S. lawmaking: legislators and agencies use generative models to summarize documents, search regulations, and draft statutory language, with examples including Representative Ted Lieu’s 2023 ChatGPT resolution and agency pilots.",{"text":82},"Capacity pressures drive adoption: South Dakota’s 70 part‑time legislators share roughly 60 staffers—the thinnest legislative staff in the country—creating demand for AI that can summarize 200‑page reports and draft amendments.",{"text":84},"Domain-specific legal copilots are emerging and scaling: platforms like Vulcan and CoCounsel integrate statutes, regulations, and citations, with some professional tools reporting about 98% accuracy on supported tasks and legal adoption rising from 14% in 2024 to 26% in 2025.",{"text":86},"Risks are concrete and measurable: the U.S. Code of Federal Regulations exceeds 200,000 pages, and unchecked AI drafting can hard‑code policy detail, reshape separation of powers, and introduce hallucinations unless disclosure, verification, and human‑review rules are required.",[88,91,94],{"question":89,"answer":90},"How are state lawmakers currently using AI to research, fact‑check, and draft legislation?","Lawmakers and staff use AI primarily for summarizing large reports, searching statutes and past committee transcripts, mapping regulatory relationships, synthesizing evidence for oversight, and drafting bill sections or amendments. In practice this means feeding past hearings, code sections, and regulatory databases into domain‑specific copilots (e.g., Vulcan, CoCounsel) to generate proposed text with citations, produce timelines, cluster facts, and flag inconsistencies. These tools shorten tasks that once took days to minutes, help nonpartisan staff and part‑time legislators keep pace with complex policy areas, and are being adopted across statehouses, federal agencies, and some foreign and local governments—while outputs still require systematic verification against primary legal sources.",{"question":92,"answer":93},"What are the main risks when legislatures rely on AI for drafting and research?","The main risks are hallucinated or misstated legal claims, omitted caveats, over‑prescriptive statutes that erode agency discretion, and reduced transparency about authorship. Faster, cheaper drafting can shift power from rulemaking agencies to legislators by hard‑coding technical details into statutes, and AI outputs can embed errors that propagate into law unless every claim and citation is checked against primary sources. These risks require procedural safeguards—mandatory disclosure of AI use, minimum human review standards, and verified citation workflows—to prevent legal and constitutional harms.",{"question":95,"answer":96},"What governance and oversight practices should statehouses adopt for AI‑assisted lawmaking?","Statehouses should require disclosure when AI substantially contributes to bill text, mandate human verification of all substantive claims and citations, vet tools for security and data governance, and provide training in lateral reading and critical verification. Agencies and legislatures should adopt standardized workflows that treat AI outputs as hypotheses, break responses into discrete claims for checking, and pair technical training with legal verification practices; early‑adopting states can pilot these frameworks and publish lessons to create transparent, accountable models for broader adoption.",[98,106,113,118,123,127,131,137,145,150,157,163,169,174,180],{"id":99,"name":100,"type":101,"confidence":102,"wikipediaUrl":103,"slug":104,"mentionCount":105},"6950428019d266277e14a252","generative models","concept",0.98,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FGenerative_model","6950428019d266277e14a252-generative-models",15,{"id":107,"name":108,"type":101,"confidence":109,"wikipediaUrl":110,"slug":111,"mentionCount":112},"69f25cb68e996ffbd5106090","Lateral reading",0.95,null,"69f25cb68e996ffbd5106090-lateral-reading",1,{"id":114,"name":115,"type":101,"confidence":102,"wikipediaUrl":116,"slug":117,"mentionCount":112},"69f25cb58e996ffbd510608a","Code of Federal Regulations","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FCode_of_Federal_Regulations","69f25cb58e996ffbd510608a-code-of-federal-regulations",{"id":119,"name":120,"type":101,"confidence":121,"wikipediaUrl":110,"slug":122,"mentionCount":112},"69f25cb58e996ffbd510608f","Regulatory operating systems",0.9,"69f25cb58e996ffbd510608f-regulatory-operating-systems",{"id":124,"name":125,"type":101,"confidence":121,"wikipediaUrl":110,"slug":126,"mentionCount":112},"69f25cb58e996ffbd510608c","Synthetic staffers","69f25cb58e996ffbd510608c-synthetic-staffers",{"id":128,"name":129,"type":101,"confidence":121,"wikipediaUrl":110,"slug":130,"mentionCount":112},"69f25cb58e996ffbd510608b","South Dakota legislature staffing","69f25cb58e996ffbd510608b-south-dakota-legislature-staffing",{"id":132,"name":133,"type":134,"confidence":135,"wikipediaUrl":110,"slug":136,"mentionCount":112},"69f25cb58e996ffbd510608d","Brazilian municipality AI-written law (2023)","event",0.86,"69f25cb58e996ffbd510608d-brazilian-municipality-ai-written-law-2023",{"id":138,"name":139,"type":140,"confidence":141,"wikipediaUrl":142,"slug":143,"mentionCount":144},"694e2fe119d266277e1497de","Virginia","location",0.99,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FVirginia","694e2fe119d266277e1497de-virginia",24,{"id":146,"name":147,"type":140,"confidence":148,"wikipediaUrl":110,"slug":149,"mentionCount":112},"69f25cb68e996ffbd5106091","Statehouses",0.88,"69f25cb68e996ffbd5106091-statehouses",{"id":151,"name":152,"type":153,"confidence":102,"wikipediaUrl":154,"slug":155,"mentionCount":156},"69398a91312dc892c4c1844b","Trump administration","organization","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FFirst_presidency_of_Donald_Trump","69398a91312dc892c4c1844b-trump-administration",178,{"id":158,"name":159,"type":153,"confidence":141,"wikipediaUrl":160,"slug":161,"mentionCount":162},"6939892b312dc892c4c18411","U.S. Department of Education","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FUnited_States_Department_of_Education","6939892b312dc892c4c18411-u-s-department-of-education",23,{"id":164,"name":165,"type":153,"confidence":166,"wikipediaUrl":167,"slug":168,"mentionCount":112},"69f25cb48e996ffbd5106089","Vulcan Technologies",0.94,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FVulcan_Centaur","69f25cb48e996ffbd5106089-vulcan-technologies",{"id":170,"name":171,"type":153,"confidence":172,"wikipediaUrl":110,"slug":173,"mentionCount":112},"69f25cb58e996ffbd510608e","Lobbyists",0.92,"69f25cb58e996ffbd510608e-lobbyists",{"id":175,"name":176,"type":177,"confidence":178,"wikipediaUrl":110,"slug":179,"mentionCount":112},"69f25cb68e996ffbd5106092","AI resolution (federal)","other",0.87,"69f25cb68e996ffbd5106092-ai-resolution-federal",{"id":181,"name":182,"type":183,"confidence":121,"wikipediaUrl":184,"slug":185,"mentionCount":186},"69f25b8d8e996ffbd510601a","Kent Roe","person","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FKent_Roe","69f25b8d8e996ffbd510601a-kent-roe",2,[188,195,202,209],{"id":189,"title":190,"slug":191,"excerpt":192,"category":11,"featuredImage":193,"publishedAt":194},"69eddbb98594a02c7d5b7537","OpenAI’s GPT-5.5: How a Unified Chat, Coding, and Browser Model Redefines Computer Work","openai-s-gpt-5-5-how-a-unified-chat-coding-and-browser-model-redefines-computer-work","1. What GPT-5.5 Is and Why It Matters\n\nGPT-5.5 is OpenAI’s newest flagship model, framed as its “smartest and most intuitive to use” and a “new class of intelligence for real work.”[1][3] It is built...","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1676272682018-b1435bad1cf0?ixid=M3w4OTczNDl8MHwxfHNlYXJjaHwxfHxvcGVuYWklMjBncHQlMjB1bmlmeWluZyUyMGNoYXRncHR8ZW58MXwwfHx8MTc3NzE5NTk2MXww&ixlib=rb-4.1.0&w=1200&h=630&fit=crop&crop=entropy&auto=format,compress&q=60","2026-04-26T09:40:11.589Z",{"id":196,"title":197,"slug":198,"excerpt":199,"category":11,"featuredImage":200,"publishedAt":201},"69ebd69aef9f887f1d4f877d","OpenAI’s GPT-5.5 Rollout: What Paid and Enterprise Users Need to Know","openai-s-gpt-5-5-rollout-what-paid-and-enterprise-users-need-to-know","OpenAI’s GPT-5.5 is framed as a “new class of intelligence for real work and powering agents,” built for complex, multi-step workflows with less user oversight.[1][3] For paid ChatGPT and Codex users,...","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1696041760711-f1bd9e111b70?ixid=M3w4OTczNDl8MHwxfHNlYXJjaHwxfHxvcGVuYWklMjByb2xsaW5nJTIwb3V0JTIwZ3B0fGVufDF8MHx8fDE3NzcwNjM1Nzh8MA&ixlib=rb-4.1.0&w=1200&h=630&fit=crop&crop=entropy&auto=format,compress&q=60","2026-04-24T20:55:57.836Z",{"id":203,"title":204,"slug":205,"excerpt":206,"category":11,"featuredImage":207,"publishedAt":208},"69e55859b951907c96a68410","GPT-Rosalind: Trusted-Access Life Sciences AI for Pharma Partners","gpt-rosalind-trusted-access-life-sciences-ai-for-pharma-partners","Pharma leaders must compress 10–15 year development timelines without compromising safety or rigor.[2][4] GPT-Rosalind, OpenAI’s new life sciences model, aims to improve the quality and speed of upstr...","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1762340275855-ae8f4c2c144e?ixid=M3w4OTczNDl8MHwxfHNlYXJjaHwxfHxncHQlMjByb3NhbGluZCUyMHRydXN0ZWQlMjBhY2Nlc3N8ZW58MXwwfHx8MTc3NjYzODA0MXww&ixlib=rb-4.1.0&w=1200&h=630&fit=crop&crop=entropy&auto=format,compress&q=60","2026-04-19T22:44:06.970Z",{"id":210,"title":211,"slug":212,"excerpt":213,"category":11,"featuredImage":214,"publishedAt":215},"69e05695e48678c58d42e3e8","How Amazon Bio Discovery Uses Agentic AI to Transform Biopharma R&D","how-amazon-bio-discovery-uses-agentic-ai-to-transform-biopharma-r-d","For biopharma leaders under pressure to cut discovery timelines and raise technical success, AI efforts often stall at proof-of-concept due to code-heavy tools and fragmented CRO workflows.[3]  \n\nAmaz...","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1632813404574-b63d317ee258?ixid=M3w4OTczNDl8MHwxfHNlYXJjaHwxfHxhbWF6b24lMjBiaW8lMjBkaXNjb3ZlcnklMjBwbGF0Zm9ybXxlbnwxfDB8fHwxNzc2MzA5OTA5fDA&ixlib=rb-4.1.0&w=1200&h=630&fit=crop&crop=entropy&auto=format,compress&q=60","2026-04-16T03:36:18.885Z",["Island",217],{"key":218,"params":219,"result":221},"ArticleBody_dsX4qbnXmGYa9QvXgZI1SNLP4kKm0E1kbBcrp53cdZg",{"props":220},"{\"articleId\":\"69f259ada569d797da77af45\"}",{"head":222},{}]