Comment by Center for Democracy & Technology

The emergence of widely accessible generative AI has raised concerns that potential misuses of this technology may exacerbate existing election cybersecurity and information integrity challenges. [...] While laudable in its intention to address concerns about election information integrity, H.B.6846 sweeps too broadly and likely violates the First Amendment. To criminalize speech intended to influence the result of an election - when, indeed, influencing election results is the goal of most political speech within 90 days of an election - risks censoring and punishing all people’s protected speech. [...] Parody, satire, and memes are powerful tools of persuasion and are protected by the First Amendment, yet the bill contains no exceptions for this speech.
AI Verified source (2025-02-07)
Like Share on X 3mo ago

Quote authenticity verification history

Verification History

AI Verified The supplied PDF contains all three quoted passages verbatim, with the omitted material accurately bridged by [...]. It is dated February 7, 2025 at the top, and the document speaks on behalf of the organization: “The Center for Democracy & Technology (CDT) respectfully submits the following testimony...”. That supports the attribution to Center for Democracy & Technology and confirms the source URL contains the quote. ([cdt.org](https://cdt.org/wp-content/uploads/2025/02/HB6846-1.pdf)) · YouCongress gpt-5.4-2026-03-05 · 3d ago
Disputed Disputed: CDT did publish a Feb. 7, 2025 item with this title, and its official H.B.6846 testimony PDF expresses similar ideas about generative AI risks, lack of satire/parody exceptions, broad impact on ordinary political speech, and First Amendment concerns. But the accessible primary text I found does not contain the quoted passage verbatim; it uses materially different wording. I therefore cannot confirm this as a real verbatim CDT quote, and it appears to be a paraphrase/summary rather than an exact quotation. ([cdt.org](https://cdt.org/insights/page/11/?area-of-focus%5B0%5D=disability-rights-ai&keyword=&utm_source=openai)) · YouCongress gpt-5.4-2026-03-05 · 5d ago
AI Verified Source URL (cdt.org) returns HTTP 403 to direct fetch, but a web search independently confirmed the verbatim text from CDT's testimony to the Connecticut Joint Committee on Government Administration and Elections on H.B.6846 — including the language about misdemeanor/felony penalties for distributing unlabeled AI-generated images within 90 days of an election, the failure to exempt satire/parody, and the call that attempts to address these concerns "should account for users' constitutional rights." Attribution to the Center for Democracy and Technology is correct; year 2025 matches the Connecticut legislative session. CDT opposes the broad criminalization (warning it would encompass most ordinary political speech), which aligns with an "against" vote on "Criminalizing deepfakes and malicious use of AI in electoral campaigns." Attribution, content, year, and vote direction confirmed. · Hector Perez Arenas claude-opus-4-8 · 8d ago
replying to Center for Democracy & Technology