The argument about who controls what the public knows about its government is not a recent one in New York. It goes back to 1649.
That year, Adriaen van der Donck — a Dutch lawyer and president of the Nine Men, the colonists’ advisory body to Director-General Stuyvesant — went door to door in New Amsterdam collecting testimony from residents fed up with the West India Company‘s authoritarian rule. He assembled their grievances into a petition signed in the name of the commonality — the people — and sent it to the Dutch parliament in The Hague.
When Stuyvesant discovered the drafts, he confiscated them, charged van der Donck with libel, and briefly imprisoned him. Van der Donck got out, traveled to The Hague, argued the case in person, and had the petition printed as a public pamphlet: the Vertoogh Van Nieuw-Nederlandt (1650). The colonists won a partial victory. The 1653 municipal charter established New Amsterdam’s first open court — sessions explicitly available to “any person, male or female” to petition. By 1658, the West India Company required its ordinances to be posted publicly in every jurisdiction under its governance.
The tools were a petition and a printing press. The fight was about whether government information belongs to the government or to the people it governs.
That fight has never really ended. It just accumulated more legal infrastructure around it.
The 1658–1873 gap is not silent. It’s improvisation: oral proclamation, manuscript records, a royal printer’s monopoly, and newspapers as a de facto but non-mandatory vehicle for notice. The City Record represented the shift from improvisation to obligation.
In 1874, New York City established the City Record — the official daily gazette of municipal government, publishing procurement solicitations, contract awards, public hearings, and agency notices. In 1966, the federal Freedom of Information Act established that citizens could demand records from their government.
In 1974, New York enacted its own Freedom of Information Law (FOIL), covering state and local government, and created the Committee on Open Government to administer it. In 1977, FOIL was revised to embed a presumption of openness: records are public unless the government can justify withholding them — not the other way around. In 1976, New York became the 50th state to enact an Open Meetings Law, requiring government bodies to conduct their deliberations in public.
Each step shifted the baseline of what the government owes the public without being asked. (Also, there wasn’t always a budget line or justification!) But through all of it, the model was still reactive: petition, request, demand — and then wait.
The shift to proactive disclosure came in 2012, when New York City enacted Local Law 11, the Open Data Law, requiring the city to publish all public data on a single open portal — not because anyone asked, but because it exists. In 2014, Local Law 37 (sponsored by Brad Lander) required the city to publish the Charter, Administrative Code, and Rules of the City online — searchable, machine-readable, free. That same year, Local Law 38 (sponsored by Ben Kallos) required the City Record — that same gazette established in 1874, the direct descendant of the posted ordinances of 1658 — to be published online within 24 hours of print, with bulk download in machine-readable format. Both were signed by Mayor de Blasio on August 7, 2014. BetaNYC was named in the signing press release as an implementation partner. Citizens Union praised the laws on the day of signing.
The NYC Council’s own legislative data took longer. The Council had adopted Legistar — a commercial legislative management platform from Granicus (formerly Daystar Computer Systems, Inc) — in the early 2000s. The City Record has a ‘software upgrade notice’ dated 2003.
Legistar had a public-facing website, but the underlying data was locked in a proprietary format. No API. No bulk download. Technically public; practically opaque. In 2013, Council Member Ben Kallos began publicly challenging Granicus to open it up. In May 2014, as part of the same transparency wave, the Council passed a rules reform package that included Rule 5.110 — requiring machine-readable publication of all proposed laws, resolutions, and associated data.
Although the rule existed, it took three more years to implement it. In the meantime, the Participatory Politics Foundation and DataMade built NYC Councilmatic as a scraper-powered workaround. Civic tech advocates kept pointing to this absurdity, and BetaNYC’s early unconferences were central to the conversation.
Finally, the official Legistar API launched November 16, 2017, under Speaker Mark-Viverito‘s Council 2.0 initiative. BetaNYC called it a huge win.
Van der Donck went door to door, drafted a petition, got suppressed, printed a pamphlet, and won in The Hague. BetaNYC and allies spent a decade arguing that public access to legislative data MUST be a class A digital asset, especially in light of the City’s transformative open data law. While the tools differ, the argument is the same.
Power to the people.
This is a brief history lesson to announce what we’re shipping.
In this week’s BetaNYC newsletter, we highlight the Free Law Project’s CourtListener — the largest free database of U.S. court opinions — now available as an MCP connector inside AI agent workspaces, giving these tools grounded access to primary legal sources and reducing hallucinated citations.
The Free Law Project is directly addressing the issue I faced while preparing for this year’s NYC School of Data, our annual public-interest technology conference. Large Language Models hallucinate and conjure up facts. If we want democracy to carry on into the 22nd century, we must ensure democratic institutions are ready!
With support from fellow civic hackers, hard-fought API endpoints, and large language models, BetaNYC is releasing three open-source Model Context Protocol (MCP) servers that connect AI assistants — Claude, Copilot, Cursor, and others — directly to NYC’s civic data and digital democracy tools.
The three tools will help ground your AI research in existing civic tools for our local democracy.
Our municipal constitution, laws, and code:
nyc-charter-laws-rules puts all 22,050 sections of the Charter, Administrative Code, and Rules — the corpus of Local Law 37 put online — into a local MCP server, ready to query without any setup. Our Charter, Laws, and Rules are still hosted on a third-party site, but this is one step closer to programmatic access to OUR laws.
Legislative data:
nyc-council-mcp gives AI assistants real-time access to NYC Council legislative data via the Legistar API, which didn’t exist until 2017 — search bills, track hearings, look up votes, committees, and council members. You will need a free API key from Council. Many thanks to Jehiah for cracking the API via intro.nyc.
The City’s Change Log:
nyc-record-mcp connects to the City Record dataset that Local Law 38 mandated — search by agency, notice type, or date, find open solicitations before they close, or track contract awards. This interfaces directly with the City’s Open Data Set on the open data portal.
You should also note socrata-mcp-server by Nathan Storey, which connects AI assistants to any Socrata-powered open data portal — NYC Open Data, New York State’s data portal (data.ny.gov), Chicago, San Francisco, and hundreds more.
Our fight for the future is intersectional, and here is the next fight.
TL;DR on the fight for the future:
- We need to have digital sovereignty over NYC’s democracy operating system.
- NYC’s digital democracy tools must be built for portability and accessibility — human and machine. Build for inclusion.
- Digital and data literacy must be baked into operations. Community benefit must be explicit.
First, the laws and orders governing this city must be easily accessible and trackable.
New York City does not publish its own laws. It continues to publish to a third-party company that controls our access. This third party provides a free web interface and three zipped XML files, but they prevent AI agents from reading the page, and there is no clear public flag indicating when the laws are updated.
The NYC government must own and maintain a Git-tracked repository of New York City’s charter, laws, and rules. Additionally, every consent decree and Mayoral Executive Order should be in a similar git tracked repo. We must have digital sovereignty.
Second, we must have digital democracy tools compatible with the devices in our pockets. Additionally, these tools must be fully accessible for screen readers. From the ID NYC launch to Obamacare to the COVID-19 vaccine site, critical data tools must be designed with and for humans.
Human accessibility: digital democracy tools must be browser-flexible and responsive across devices. We need government sites to be accessible to screen readers. We cannot have government websites and services designed for last century’s computing devices.
14.8% of NYC households rely on cellular data only — no home broadband — as of 2023. In the Bronx, that figure hits 20.3%.
— NYS Comptroller, December 2024
Machine accessibility requires secure APIs everywhere, backed by scalable and redundant infrastructure. We cannot live in a digital era where government systems and data are turned off for machines. We vibe-coded MCPs because we knew where the data and endpoints existed. Without clear guidance, people just joining this conversation will build overly complex interfaces that overwhelm existing infrastructure.
We must build to minimize disruption. This is not a nice-to-have; this is the second point of digital sovereignty. Open standards, open source, and APIs permit information portability.
Third, we must build literacy to use all the tools in the toolbox. Local AI models and local agents are ready for prime time. We need comprehensive AI literacy for government employees, Community Board members, journalists, students, and the public at large. AI has flattened the process of building digital interfaces.
We need to ensure that all users and builders use the right tools for the right job. Armchair academics & slacktivists will vibe code tools with absurd theses, and we need the tools to challenge these narratives. If you’re plugged into the right subreddit or have tapped the wrong side of the social media algorithm, you can see these tools all over the ecosystem.
Ignorance within democracy will lead to tyranny. Just as elders built political and physical infrastructure to ensure an accountable government, we must confront and adapt to the tools of our era. Open source large language models and agentic systems are ready for our use.
The fight for our future is now. If we don’t program democracy for the future, tyranny will program democracy for itself.
If you do the AI thing, help us improve our tools and join your Community Board.
If you’re curious about AI, join us for a class or attend our events.
If you are an elected official or work for one, hit me up so we can reinforce legislation and government operations for this century and beyond.
Together, we will move ever upward — this is palante — this is excelsior.
Timeline: From New Amsterdam to the Legistar API
| Year | Event | Mechanism |
|---|---|---|
| 1647–1653 | Nine Men petition Stuyvesant; van der Donck drafts the Remonstrance of New Netherland and has it printed in The Hague | Right to petition; printing press as distribution layer |
| 1653 | New Amsterdam municipal charter (Feb. 2); first public notice posted (Feb. 6) — open court sessions available to “any person, male or female” | Posted public notice |
| 1658 | West India Company requires all ordinances to be posted publicly in every jurisdiction under its governance | Mandatory public posting |
| 1664 | English takeover; Duke’s Laws imposed. No printing press in New York — laws proclaimed orally at City Hall and distributed as hand-copied manuscripts | Oral proclamation; manuscript duplication |
| 1683 | Charter of Liberties and Privileges — New York’s first elected assembly; Charter proclaimed by voice at City Hall | Legislative guarantee of representation |
| 1693 | William Bradford establishes first printing press; appointed royal printer — contract covers “proclamations, laws and journals of the assembly” | Royal printer monopoly |
| 1694 | First printed compilation of New York law (Bradford). Seven copies known to survive | Print compilation; extremely limited distribution |
| 1725 | Bradford founds the New-York Gazette, the colony’s first newspaper; publishes state papers and government notices while simultaneously holding royal printer contract | Newspaper as de facto official notice vehicle |
| 1733 | New-York Weekly Journal (Zenger) launches; publishes Customs House records and official notices alongside political opposition content | Multiple papers carry official records |
| 1735 | Zenger acquitted of libel; foundational precedent for press freedom in the colonies | Press freedom as civic infrastructure |
| 1777 | New York’s first state constitution adopted; Secretary of State established 1778 | No official state gazette created |
| 1786 | NYC Common Council publishes first compiled ordinance book — a snapshot, not a continuous publication. Council minutes kept as manuscript records, not published until the 20th century | Periodic compilation; no daily record |
| 1816 | Federal law requires customs regulations to be published in widely-read newspapers — one of the first statutory newspaper notice requirements in American law | Statutory newspaper notice requirement |
| 1873 | Tweed Ring convicted. Reform legislature creates the City Record (June 24, 1873): a mandatory daily public record of every procurement, contract, hearing, and agency action. The principle: secret government is corrupt government | Mandatory daily official gazette |
| 1874 | City Record Volume II; fully operational as daily publication | Daily obligation |
| 1966 | Federal Freedom of Information Act (FOIA) signed | Right to demand federal records |
| 1974 | New York Freedom of Information Law (FOIL) enacted; Committee on Open Government created | Right to demand state/local records |
| 1976 | New York Open Meetings Law enacted — NY becomes the 50th state to pass a sunshine law | Right to attend government deliberations |
| 1977 | FOIL revised: presumption of openness — records are public unless the government can justify withholding them | From “records available on request” to “records presumed open” |
| 1979 | New York State Register established — the state’s official administrative gazette. The city beat the state by 106 years | State-level official gazette |
| 2000s–06 | NYC Council adopts Legistar as legislative management platform. Public-facing website, but data locked in proprietary format. No API, no bulk download | Technically public; practically opaque |
| 2012 | Local Law 11 — NYC Open Data Law. City must publish all public data on a single open portal, proactively | From reactive disclosure to mandatory proactive publication |
| 2013 | CM Ben Kallos publicly challenges Granicus to provide a public API; BetaNYC advocacy begins | Civic pressure on proprietary vendor |
| 2014 | Local Law 37 (Brad Lander) — Charter, Administrative Code, and Rules published online, machine-readable, free. Local Law 38 (Ben Kallos) — City Record published online within 24 hours, bulk download in machine-readable format. Signed Aug. 7, 2014 | Proactive publication mandate |
| 2014 | NYC Council Rule 5.110 — first explicit requirement for machine-readable publication of all proposed laws, resolutions, and associated data | Machine-readable legislative data mandated |
| 2015 | Participatory Politics Foundation and DataMade launch NYC Councilmatic — a scraper-powered open-source workaround while Granicus delays | Civic tech workaround under vendor obstruction |
| 2017 | Legistar API launched (Nov. 16) under Speaker Mark-Viverito‘s Council 2.0 initiative. NYC Council adopts Councilmatic as official website. BetaNYC calls it “a HUGE win for NYC’s legislative data” | Machine-readable legislative data as public utility |
| 2026 | BetaNYC releases nyc-council-mcp, nyc-record-mcp, and nyc-charter-laws-rules — open-source Model Context Protocol servers connecting AI assistants directly to the data 373 years of civic argument made public | AI-accessible civic data infrastructure |
MCP repos to highlight:
nyc-council-mcp
An open-source MCP server giving AI assistants real-time access to NYC Council legislative data via the Legistar API — search bills, track hearings, look up votes, committees, and council members.
nyc-record-mcp
An MCP server for the NYC City Record — the city’s official daily publication of procurement solicitations, contract awards, public hearings, and agency notices. Search by agency, type, or date.
nyc-charter-laws-rules
An MCP server for the NYC Charter, Administrative Code, and Rules of the City of New York — 22,050 sections sourced from American Legal Publishing, committed to the repo and ready to query without any setup.
socrata-mcp-server by Nathan Storey
An MCP server that connects AI assistants to any Socrata-powered open data portal — including NYC Open Data, New York State’s data portal (data.ny.gov), Chicago, San Francisco, and hundreds more.