46% of the ChatGPT-named Telegram channels we could date predate its public debut

A Telegram channel can outlive the name on its door. Of 620 indexed public channels whose current title mentions ChatGPT, we could date 117 from their numeric IDs. For 54 of those, the entire estimated creation interval ends before OpenAI's November 30, 2022 public introduction: 46.2% of the dated subset. That is evidence of older channels adopting a newer name, subject to the dating model and naming assumptions.

An old ID beneath a new title

OpenAI's dated introduction of ChatGPT gives us a useful historical boundary. For example, our snapshot lists @ultra_ai as “ChatGPT | Midjourney Bot”. Its ID maps to September 2015–January 2017. The channel itself appears much older than the public product name. We do not know its earlier title, topic, owner, or subscriber count.

Our channel ID study maps IDs to creation windows. We call a channel older only when the whole window ends before the naming cutoff, with a seven-day guard. Overlapping windows stay uncertain. A newer window only dates the channel after the public record; it does not establish that the channel has never changed its name.

Larger channels show the same pattern

Missing IDs concentrate among small channels, so we repeated the count above subscriber thresholds. At 1,000 subscribers, 46 of 98 dated matches are older (46.9%); at 10,000, 31 of 58 are older (53.4%). The pattern survives better ID coverage, although neither subset is a random sample of Telegram.

The 54 older ChatGPT channels have a median of 13,153 subscribers, compared with 8,291.5 for the 40 newer channels; subscriber counts are known for all 94. This is a descriptive gap. Older channels have had more time to grow, and we observe survivors. It cannot show that renaming caused audience growth or that the existing audience followed the previous topic.

Two dated names, three exploratory cutoffs

DALL·E has another clear official naming record: OpenAI's January 5, 2021 announcement. None of its six dated matches is older. The full cohort has only 11 channels and excludes unpunctuated “dalle”, a common Italian word. A small null result does not establish an absence of renaming.

For three other names, the historical boundary is less secure. These are conditional exploratory results, not equivalent proof of first public naming. DeepSeek uses May 1, 2023, ahead of a May 25 interview describing the Chinese organization's naming that month; the first English use remains unresolved. Notcoin uses October 1, 2023, before its reported November beta (a TON Foundation article preserved in a third-party archive). Hamster Kombat uses January 1, 2024, before its surviving March 14 title-change record. October 1 and January 1 are chosen buffers, not discovered launch dates.

These yield 5 older channels out of 8 dated DeepSeek matches, 12 of 32 for Notcoin, and 10 of 22 for Hamster Kombat. Earlier private communities or earlier uses of the same names remain possible. Cohorts overlap, so their rows must not be added to count distinct renamed channels. We excluded generic names, predictable sequels such as GPT-4, and release dates that followed earlier announcements.

Current metadata cannot reconstruct a previous identity

Among the 54 older ChatGPT channels, current language metadata identifies 31 as Russian, 11 as English, two each as Ukrainian, Persian and Spanish, and one as Arabic; five lack a language label. Only 25 have mapped topic metadata: 24 carry Technology & IT and six Education & science. Topic families overlap; 29 channels lack mapped topics. Among the 40 newer matches, 34 have language labels (20 Russian, nine English) and 17 have mapped topics (16 Technology & IT, three Education & science). These asynchronous saved labels describe observed languages and themes, not previous identities.

Nothing here establishes a channel sale, a change of owner, deception, or a previous subject. A publisher could have renamed an existing technology channel as its coverage evolved. The measurable finding is the mismatch between a current name and an estimated creation period; identifying the actual transition requires historical titles or posts.

A reproducible test, with a deliberately narrow claim

We searched current titles only, using case-insensitive patterns; usernames and descriptions did not select candidates. ChatGPT permits a space or hyphen between Chat and GPT, while GPT alone does not qualify. We kept records marked as existing, explicitly non-group channels, deduplicated by normalized peer ID (falling back to username for missing IDs), and retained the latest checked record. The underlying index contains 3,051,602 eligible public channels. Its observations span April 14–October 10, 2026; October 10 is the latest check, not a simultaneous census or this article's publication date.

The surviving public-channel index misses private, deleted and unindexed channels. IDs and metadata are unevenly available; threshold comparisons omit unknown subscriber counts. Our observations cannot locate the rename in time or measure how frequently all Telegram publishers rename. They do show why a channel's apparent age needs two questions: when was the channel created, and when did its present name become possible?

Download the aggregate JSON, matching patterns and source register. Check an individual ID with the creation-date tool, read the ID methodology, or explore channels with matching names.

Suggested citation: TGScope Research. “46% of the ChatGPT-named Telegram channels we could date predate its public debut.” October 17, 2026. Aggregate snapshot: October 10, 2026.