Good morning, Council Members. My name is Ben Weinberg, and I am the Director of Public Policy at Citizens Union, a nonpartisan good government group that has worked for honest and accountable government and open and fair elections in New York for nearly 130 years.

We thank the Council for holding this important hearing. We will comment on two bills before you today, both of which tackle the democratic challenges faced by emerging technologies: Introduction 12-A-2026 from Speaker Julie Menin, which addresses AI-generated content in local elections, and Introduction 951-2026, sponsored by Council Member Shanel Thomas-Henry, which focuses on public servants’ use of prediction markets.

Introduction 951-2026: Preventing Policymakers and Their Relatives from Trading on Prediction Markets Using Insider Information

Prediction markets present new risks for public ethics. An industry that allows betting on the outcomes of real-world future events creates opportunities for individuals who influence such events—or who are privy to the planning of such events—to abuse their privileged access for private gain. In several prominent recent examples, government officials used their special access to place bets on events they knew would occur, including military actions and executive decisions,¹ cashing in big sums of money.

We commend Council Member Thomas-Henry for tackling this emerging field, and offer the offer the following comments on the proposed legislation.

• Address this issue through the City’s existing conflicts of interest law

Much of the conduct targeted by Intro 951-2026 is already forbidden under the City’s conflicts of interest law, City Charter Chapter 68. If a New York City public servant had engaged in prediction markets trading of the type described in the examples above, they would have likely been investigated and penalized by the Conflicts of Interest Board (COIB).

Section 2604(b)(3) of the Charter prohibits public servants from using their position to obtain financial gain, and Section 2604(b)(4) prohibits public servants from disclosing “any confidential information concerning the property, affairs or government of the city which is obtained as a result of the official duties of such public servant and which is not otherwise available to the public” to advance their financial interests.

Importantly, those sections are broader than the proposed bill because they apply to all public servants, not only to individuals on the “policymaker” list.

To the extent that new risks posed by prediction markets are not covered by existing prohibitions, targeted additions could be incorporated into Section 2604 of Chapter 68, which lists prohibited interests and conduct.

That would make effective use of existing definitions, enforcement processes, and penalty levels, thus reducing confusion created by having a separate law and streamlining oversight. For example, Chapter 68’s definitions for “confidential information,” relatives, associates, and policymakers have already been interpreted in advisory opinions and applied in prior enforcement actions, making them useful regulatory tools in these cases.

• Do not apply the bill and its associated penalties to relatives of public servants

The obligation to protect the integrity of public office rests on the public servant, not their spouses. Expanding COIB’s enforcement jurisdiction to public servants’ spouses, parents, or siblings would be legally fraught and difficult to implement.

• Consider alternative approaches if council members find that existing prohibitions are insufficient to preserve public integrity around prediction markets betting:
  • Ban on job-related prediction market bets: Prohibit public servants from trading on prediction market events where the subject matter relates to their official duties or their agency’s decisions, regardless of whether they had direct involvement in a specific decision or information about the event’s outcome. An executive order issued last month by North Carolina Governor Josh Stein has taken that approach.
  • Full ban on prediction market bets: The U.S. Senate has imposed the most aggressive restriction thus far, banning its members and staff from placing any bets on prediction markets. Lawmakers in several states have proposed similar measures, but such proposals raise legal questions and enforcement challenges.
  • Tailored definition of “nonpublic information”: In consultation with the Conflicts of Interest Board, lawmakers can examine whether the current interpretation of “confidential information,” as defined in Chapter 68, adequately addresses all relevant prediction market risks, and if not, whether a tailored definition that applies to the use of other types of information should be added. Governor Hochul’s most recent Executive Order on prediction markets is grounded on the existing State Code of Ethics’ rules on using “confidential information” but adds new restrictions that apply to using “nonpublic information obtained in the course of [public servants’] official duties”.
  • Require written affirmation of existing restrictions: Administratively, public servants could be required to sign an acknowledgment that they understand and have complied with existing prohibitions on trading confidential information for private gain, including in prediction markets. Signing such a document carries legal weight, as a person may be charged with a misdemeanor for knowingly filing a false affirmation.

Introduction 12-A-2026: Requiring Disclosure of Digitally Manipulated Content in Local Elections

Another emerging technology that carries democratic risks is artificial intelligence, and specifically the use of AI-generated content in elections. We thank Speaker Menin for confronting this topic.

Intro 12-A-2026 would require campaigns and independent expenditure committees to submit provenance data on AI-generated content to the Campaign Finance Board (CFB).

A key part of combating deepfakes in campaigns is ensuring voters know that the ad they are seeing or hearing was generated by AI. The most direct way to accomplish that is by requiring a disclosure on the ad itself – similar to the “paid for by” attribution already required under campaign finance law.

Such disclosure is already mandated by state law. New York Election Law Section 14-106 currently requires any campaign or committee that distributes political communications containing “materially deceptive media” to disclose that use to voters.

This AI disclosure requirement, signed into law in 2024, presumably applies to local candidates and independent expenditure committees active in New York City elections. Enforcement of election law violations is under the jurisdiction of the State Board of Election and the Division of Election Law Enforcement, where investigative actions are often delayed or paused. However, the law allows candidates whose likeness was generated by AI to petition the court to force enforcement.

Intro 12-A-2026 would complement the existing state framework by creating a public paper trail of any use of AI in campaign material. The CFB will be the holder of such information. We note that under current city law, campaigns are not required to submit advertisements to the CFB – only independent expenditure committees are subject to that requirement.

We also flag one key technical point: The bill references state law provisions that had not yet been enacted. The specific statutory references may need to be updated as the relevant state law develops.

Thank you for the opportunity to testify. I am happy to answer any questions.