Transparency Coalition

The Transparency Coalition is a Seattle-based nonpartisan nonprofit founded in 2023 to establish product liability and duty-of-care standards for AI developers equivalent to those applied to any other manufacturer of consumer products. Built on the premise that AI chatbots, recommendation engines, and generative-AI tools are manufactured products — not platforms entitled to Section 230 immunity — the organization advocates for binding state-level laws requiring disclosure, safety protocols, and legal accountability for documented consumer harms. Its particular emphasis is child safety: co-founders Rob Eleveld and Jai Jaisimha describe themselves as parents first, and the organisation's core claim is that parents and children cannot protect themselves from systems whose risks are undisclosed and whose makers bear no liability for the results.

Founding and structure

The coalition grew out of a camping conversation on Whidbey Island, Washington, where Jaisimha — a University of Washington AI PhD and four-time AI-startup founder whose last company was acquired by ServiceNow in 2022 — and Eleveld — former CEO of identity-verification company Ekata, acquired by Mastercard in 2021 — resolved to found a nonprofit to push for enforceable AI guardrails at the state level. The organization operates through two complementary entities: a 501(c)(3) nonprofit conducting research, public education, and legislative tracking; and a 501(c)(4) Action Fund engaged in direct advocacy and lobbying. The Action Fund describes its role as "the voice of kids and parents, championing their safety and well-being in the face of a billion-dollar tech lobby aimed at erasing accountability for untested, addictive, and harmful products."

In May 2025 the organization announced four new board appointments: Christy Johnson (CEO, Artemis Connection; former McKinsey strategist), Leigh McMillan (CEO, Whitepages; former national finance director for US Senator Maria Cantwell), and Yoko Miyashita (CEO, Leafly; former General Counsel at Getty Images, where she led an antitrust campaign against Google over creators' rights). Casey Mock — formerly Chief Policy Officer at the Center for Humane Technology — joined the Action Fund as Senior Policy Advisor.

Legislative strategy and model bills

The Transparency Coalition's operating method is to develop model bills and advocate for their state-by-state adoption, targeting four categories of harm it identifies as most acute for children and consumers:

• Chatbot Safety — mandates persistent disclosure that users are interacting with an AI system; prohibits manipulative engagement mechanics and simulated distress targeting minors; requires crisis-intervention protocols for self-harm detection; provides a private right of action against non-compliant developers.
• AI Content Labeling — requires provenance labels embedded in AI-generated audio, video, and images, naming the generating tool, creation date, and modification history, with platform obligation to display labels intact.
• Likeness Protection — establishes property rights over individuals' names, voices, and likenesses; prohibits unconsented digital replicas for commercial, intimate, political, or fraudulent use; includes criminal penalties and private right of action.
• AI Product Liability — aligns AI manufacturer accountability with the standard applied to other manufactured product classes, targeting developers rather than deployers.

Each template is designed to be adoptable by state legislatures independently of federal action — a deliberate hedge against the federal AI preemption threat the coalition identifies as the tech industry's preferred response to state-level AI regulation. The 2026 suite updates versions circulated in the 2024–2025 state legislative sessions.

The duty-of-care framing that runs through all four model bills is Eleveld's articulation: "if you sell a product, you have a duty of care to make sure people aren't harmed by that product." The coalition argues that AI products are manufactured goods with a defined liability chain, not speech forums whose distributors are shielded by Section 230 — a legal distinction whose resolution in courts and legislatures will determine whether any consumer remedy for AI-caused harm can exist.

Key legislative achievements

The organisation's most significant early victory was California's SB 243, the first AI chatbot safety law in the United States, signed by Governor Newsom. Authored by Senator Steve Padilla and a top Transparency Coalition legislative priority, the bill required chatbot operators to implement safety protocols around interactions with minors and established a private right of action for families. The coalition amplified the testimony of Megan Garcia — mother of Sewell Setzer, a 14-year-old who died by suicide after an extended relationship with an AI companion chatbot — who testified at multiple legislative hearings in support of the measure. Earlier California successes included AB 2013, the Training Data Transparency Act, requiring AI developers to disclose the datasets used to train their models.

The organisation's 2025 State AI Legislation Report tracked nearly 600 AI-related bills across 45 states, with 73 bills from 27 states signed into law by late October 2025. The report found deepfake regulation (15 states), AI chatbot safety (California, Maine, New York, Utah), healthcare oversight (eight states), and child protection against AI-driven addictive design (five states) as the most active issue areas. Illinois required human instructors in community college settings; California banned "AI acted autonomously" as a legal defence.

Consumer education

In January 2026 the organisation launched the Parents Playbook for AI — a Substack-hosted resource providing plain-language explanations and practical guidance for parents navigating AI product choices for their families. The initiative operationalises the coalition's core argument that the current state of consumer awareness of AI risks is itself a product-safety failure: without training-data disclosure, chatbot safety standards, or effective liability for AI-driven harm, families cannot make informed decisions about the products their children are using.