conv.

All stories
AIQuiet 15d · day 17

UK watchdog weighs probe into Home Office's OpenAI-based asylum tools

A government inspector is considering investigating the Home Office's ChatGPT-based tools after a digital rights group found error rates in asylum case summaries higher than officially reported.

What to know

  • ICIBI, the UK's borders and immigration watchdog, is weighing an inspection of the Home Office's OpenAI-based asylum tools after requesting more information from Open Rights Group.
  • ORG's six-month investigation found the error rate in AI-generated case summaries is significantly higher than the Home Office's published 9 percent pilot figure, and that one of the tools has no quality control.
  • The Home Office has withheld details on the tools despite 11 parliamentary questions, two ministerial letters and 10 public records requests, and maintains the AI only assists caseworkers who make the final decisions.
  • The tools, used since late 2024 to help clear a roughly 100,000-case asylum backlog, carry high stakes since decisions affect whether applicants are returned to countries where they fear persecution.

“effectiveness and reliability were varied”

ICIBI, UK government watchdog · WIRED ↗ · Oct 26

Sara Alsherif Open Rights Group campaignerOpen Rights Group Digital rights nonprofitIndependent Chief Inspector of Borders and Immigration (ICIBI) Government watchdogUK Home Office Government department using the tools

UK watchdog weighs probe into Home Office's OpenAI-based asylum tools
wired.com

How it unfolded 2 developments, newest first · click a bar or a number to jump articlesposts

Peak 3 pieces in 4h at Oct 27, 11 AM; 6 pieces over 18 days (3 articles · 3 posts) Oct 27, 11 AM — 3 pieces · 3 articles — Google News 1, Mastodon 1, Newswires 1Oct 27, 3 PM — quietOct 27, 7 PM — quietOct 27, 11 PM — quietOct 28, 3 AM — 2 pieces · 2 posts — Mastodon 1, Reddit 1Oct 28, 7 AM — quietOct 28, 11 AM — quietOct 28, 3 PM — quietOct 28, 7 PM — quietOct 28, 11 PM — quietOct 29, 3 AM — quietOct 29, 7 AM — quietOct 29, 11 AM — quietOct 29, 3 PM — quietOct 29, 7 PM — quietOct 29, 11 PM — quietOct 30, 3 AM — 1 piece · 1 post — Mastodon 1Oct 30, 7 AM — quietOct 30, 11 AM — quietOct 30, 3 PM — quietOct 30, 7 PM — quietOct 30, 11 PM — quietOct 31, 3 AM — quietOct 31, 7 AM — quietOct 31, 11 AM — quietOct 31, 3 PM — quietOct 31, 7 PM — quietOct 31, 11 PM — quietNov 1, 2 AM — quietNov 1, 6 AM — quietNov 1, 10 AM — quietNov 1, 2 PM — quietNov 1, 6 PM — quietNov 1, 10 PM — quietNov 2, 2 AM — quietNov 2, 6 AM — quietNov 2, 10 AM — quietNov 2, 2 PM — quietNov 2, 6 PM — quietNov 2, 10 PM — quietNov 3, 2 AM — quietNov 3, 6 AM — quietNov 3, 10 AM — quietNov 3, 2 PM — quietNov 3, 6 PM — quietNov 3, 10 PM — quietNov 4, 2 AM — quietNov 4, 6 AM — quietNov 4, 10 AM — quietNov 4, 2 PM — quietNov 4, 6 PM — quietNov 4, 10 PM — quietNov 5, 2 AM — quietNov 5, 6 AM — quietNov 5, 10 AM — quietNov 5, 2 PM — quietNov 5, 6 PM — quietNov 5, 10 PM — quietNov 6, 2 AM — quietNov 6, 6 AM — quietNov 6, 10 AM — quietNov 6, 2 PM — quietNov 6, 6 PM — quietNov 6, 10 PM — quietNov 7, 2 AM — quietNov 7, 6 AM — quietNov 7, 10 AM — quietNov 7, 2 PM — quietNov 7, 6 PM — quietNov 7, 10 PM — quietNov 8, 2 AM — quietNov 8, 6 AM — quietNov 8, 10 AM — quietNov 8, 2 PM — quietNov 8, 6 PM — quietNov 8, 10 PM — quietNov 9, 2 AM — quietNov 9, 6 AM — quietNov 9, 10 AM — quietNov 9, 2 PM — quietNov 9, 6 PM — quietNov 9, 10 PM — quietNov 10, 2 AM — quietNov 10, 6 AM — quietNov 10, 10 AM — quietNov 10, 2 PM — quietNov 10, 6 PM — quietNov 10, 10 PM — quietNov 11, 2 AM — quietNov 11, 6 AM — quietNov 11, 10 AM — quietNov 11, 2 PM — quietNov 11, 6 PM — quietNov 11, 10 PM — quietYesterday, 2 AM — quietYesterday, 6 AM — quietYesterday, 10 AM — quietYesterday, 2 PM — quietYesterday, 6 PM — quietYesterday, 10 PM — quietToday, 2 AM — quietToday, 6 AM — quietToday, 10 AM — quietToday, 2 PM — quietToday, 6 PM — quiet 12
Oct 28Oct 30Nov 1Nov 3Nov 5Nov 7Nov 9Nov 11now · 10:09 PM ET
  1. 2

    Open Rights Group publicizes probe and its findings

    ORG posts on Mastodon highlighting the possible Home Office probe and restating that its investigation found errors and inaccuracies in the ChatGPT-based tools used to summarize interviews and policy documents.

    “The UK Home Office may face a probe by a government watchdog 🚨 This comes following the findings in ORG's six-month investigation into the use of AI in asylum decision-making. We found errors and inaccuracies in the ChatGPT tools being used to summarise interviews and policy documents.”
    — Open Rights Group
    • openrightsgroup@social.openrightsgroup.org

      The UK Home Office may face a probe by a government watchdog 🚨 This comes following the findings in ORG's six-month investigation into the use of AI in asylum decision-making. We found errors and inaccuracies in the ChatGPT tools being used to summarise interviews and policy documents. Read more ⤵️ https://www. wired.com/story/uk-home-office…

      openrightsgroup@social.openrightsgroup.orgMastodon14d ago30▲view on Mastodon ↗
  2. 2 days quiet
  3. 1

    ICIBI weighs inspection of Home Office's AI tools, Wired reports

    WIRED reports the Independent Chief Inspector of Borders and Immigration has asked ORG for more information on the tools as it plans next year's inspection topics; the Home Office defends the tools as assistive only, not decision-making.

    “Asylum decisions are made by trained caseworkers, with AI tools used only to assist in navigating interview records and policy information. They cannot be used in isolation and do not make decisions on claims…”
    — Home Office spokesperson
    1. first by Mastodon, 17d ago · also Wired

      1 more headline
    • clarinette@mastodon.online

      https://www. wired.com/story/uk-home-office -may-face-investigation-over-use-of-openai-tools/

      clarinette@mastodon.onlineMastodon16d agoview on Mastodon ↗
  4. background

    Open Rights Group completes six-month probe into the tools — ORG's investigation, involving 11 parliamentary questions, two ministerial letters and 10 public records requests, finds the share of ACS summaries deemed not fit for use is significantly higher than the published 9 percent pilot error rate, and that APS summaries have no quality control.

  5. background

    Home Office begins piloting OpenAI-based asylum tools — The Home Office starts piloting the Asylum Case Summarisation (ACS) and Asylum Policy Search (APS) tools, built on OpenAI's 2023 GPT-4 model and Microsoft Foundry, to help clear a backlog of around 100,000 asylum cases.

What people are saying 0 voices from 0 sites · best of 2 · verbatim