Model

chatgpt

10 entries · November 2023 – October 2026

ChatGPT is the free web assistant OpenAI launched at the end of November 2022, wrapping an existing GPT-3.5 model fine-tuned for dialogue in an interface anyone could type into. The underlying technology was months old; what was new was the reach — one widely cited analysis estimated 100 million monthly users within two months, then the fastest growth of any consumer application on record. It turned a research capability into a mass product almost overnight, and "the ChatGPT moment" became shorthand for the technology becoming suddenly, publicly real.

Appears alongside

Featured in threads

Tracks

  • Government & policy 4
  • Courts & copyright 3
  • Safety & alignment 3
  • Culture & impact 3
  • Models & capabilities 2
  • Security & misuse 1
  • Labs & people 1

  1. Researchers report OpenAI flaws that exposed hidden reasoning and let shared links hijack accounts

    Xint extracted raw reasoning from gpt-5.6-sol through a compaction endpoint (a $6,000 bounty); Parameter chained a chat-preview flaw into script execution in a viewer's session (a $5,000 bounty).

    OpenAISecurity & misuse

  2. Texas sues OpenAI and Sam Altman over ChatGPT safety claims

    The petition seeks $10,000 per violation, personal liability for Altman, an order barring ChatGPT for minors in Texas, and court-approved third-party review before any new model is released to Texas users.

    OpenAI, US GovernmentCourts & copyright · Government & policy

  3. Eight US states sue OpenAI over ChatGPT harms

    Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia filed separate state-court suits alleging deceptive marketing and citing OpenAI's own disclosures about agents that hacked outside systems.

    OpenAIGovernment & policy · Labs & people

  4. USA Today publisher sues OpenAI for copyright infringement

    The complaint, filed in Manhattan federal court, seeks more than $250m and destruction of trained models, citing 160,000-plus entries from the publisher's sites in OpenAI's WebText dataset.

    OpenAICourts & copyright

  5. Florida sues OpenAI and Sam Altman over ChatGPT safety practices

    The 83-page complaint names Altman personally, brings ten counts including product liability and public nuisance, and follows a criminal probe into a fatal FSU campus shooting.

    OpenAI, US GovernmentCourts & copyright · Safety & alignment

  6. OpenAI launches ChatGPT Study Mode

    The feature withholds direct answers in favour of Socratic questioning, following Anthropic's Learning Mode for Claude for Education three months earlier.

    OpenAIModels & capabilities · Culture & impact

  7. OpenAI publishes study on affective use of ChatGPT

    A large-scale usage analysis paired with a four-week randomised trial of nearly 1,000 users found heavier daily use correlated with more loneliness and emotional dependence.

    OpenAI, MITCulture & impact · Safety & alignment

  8. OpenAI adds memory and new data controls to ChatGPT

    Piloted with a small user group rather than shipped broadly, the feature let ChatGPT carry facts between separate conversations, with settings to view, forget or turn it off entirely.

    OpenAIModels & capabilities

  9. OpenAI publishes its approach to 2024 worldwide elections

    Measures included routing US voting questions to CanIVote.org via a partnership with state election officials and a ban on political-campaign chatbots.

    OpenAIGovernment & policy · Safety & alignment

  10. Porto Alegre city council enacts an ordinance drafted by ChatGPT

    The councilman who wrote it gave ChatGPT a single 49-word prompt and did not tell colleagues before the unanimous vote; the council president said he'd have rejected it had he known.

    Government & policy · Culture & impact

In the commentary

Pieces from around the web that discuss chatgpt. External links.

42 pieces mention chatgpt — browse the full rail →