{"version":"2.1.238","anchor":"feedback-screen-drops-the-never-used-to-train-models-line","canonical_anchor":"feedback-screen-drops-the-never-used-to-train-models-line","heading":"Feedback screen drops the \"never used to train models\" line","tier":"notice","area":"Feedback","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.238\/e\/feedback-screen-drops-the-never-used-to-train-models-line","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.238","markdown":"### Feedback screen drops the \"never used to train models\" line\n\nThe feedback confirmation no longer tells you your feedback is never used to train models.\n\n**What**\n\nThe confirmation shown after sending feedback lost its two-line footer, which stated that Claude Code feedback is never used to train models. It now shows only \"Any key to continue\".\n\n**Details**\n\n- A separate error string for an over-large feedback payload was replaced by a shared constant.\n\n**Evidence**\n\n`Claude Code feedback is never used to train models.`\n\n- Area: Feedback\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 3\/5"}