Checked Nov 5, 2026Start a new check

What was checked

ChatGPT constantly recommends a fake crypto scam clone site (t33.codes, registered 2 weeks ago) when users ask for the real T3 Code download link (t3.codes), making AI search a new phishing vector.

Assessment summary

Needs caution

The reported incident could illustrate a phishing risk involving AI-generated recommendations. But one allegation does not establish that AI search broadly constitutes a new phishing vector. ChatGPT does recommend a lookalike domain (t33.codes) over the real t3.codes, but sources describe it as a copied website competing in search rankings, not specifically as a 'crypto scam.' T3 Code is a coding agent interface, not a cryptocurrency product. The 'crypto scam' characterization appears inaccurate.

Key findings

  • ChatGPT constantly recommends a fake crypto scam clone site (t33.codes)

    Needs context
    ChatGPT does recommend a lookalike domain (t33.codes) over the real t3.codes, but sources describe it as a copied website competing in search rankings, not specifically as a 'crypto scam.' T3 Code is a coding agent interface, not a cryptocurrency product. The 'crypto scam' characterization appears inaccurate.
    Perplexity:
    A secondary report says users received t33.codes from ChatGPT and calls it a scam clone. However, the available sources do not independently establish that ChatGPT does this constantly, that t33.codes is definitively fraudulent, or that it is a crypto scam.
  • This behaviour makes AI search a new phishing vector

    ClaudeSupported
    Multiple 2026 sources document cloned retail sites appearing in ChatGPT results, fake AI ad portals, and fraudulent websites being recommended by AI search tools, confirming this is an active threat vector.

    No linked source for this finding.

    PerplexityNeeds context
    The reported incident could illustrate a phishing risk involving AI-generated recommendations. But one allegation does not establish that AI search broadly constitutes a new phishing vector.
    ChatGPTCan’t verify
    This single reported incident shows a risk that AI outputs can surface malicious domains, but it does not by itself prove AI search is an established, systematic phishing vector across products and time.
  • ChatGPT constantly recommends a fake crypto scam clone site (t33.codes) when users ask for the real T3 Code download link (t3.codes)

    Needs context
    Multiple social-media posts report ChatGPT suggesting t33.codes for T3 Code downloads, but 'constantly' overstates the available evidence and independent replication of frequency is lacking.
  • t33.codes was registered 2 weeks ago

    Can’t verify
    The available search results repeat the two-week registration claim but provide no WHOIS, registry, or other primary evidence confirming the date.
    ChatGPT:
    Public posts and reporting around Nov 3–4, 2026 state the t33.codes domain was registered around Oct 19, 2026, which is about 2–2.5 weeks before Nov 5, 2026.

Model assessments

Assessment summary

  • The real T3 Code site and download instructions are corroborated by the official site and related GitHub repositories t3.codes t3.codes github.com.
  • A secondary source repeats the allegation that ChatGPT recommended t33.codes and describes it as a clone registered two weeks earlier, but the available evidence does not independently verify ChatGPT’s repeated behavior, the registration date, or that the site is definitively a scam zeronoise.ai.

Claims reviewed

This makes AI search a new phishing vector

Needs context
The reported incident could illustrate a phishing risk involving AI-generated recommendations zeronoise.ai. But one allegation does not establish that AI search broadly constitutes a new phishing vector.

This claim may mislead readers without added context.

ChatGPT constantly recommends a fake crypto scam clone site (t33.codes)

Can’t verify
A secondary report says users received t33.codes from ChatGPT and calls it a scam clone zeronoise.ai. However, the available sources do not independently establish that ChatGPT does this constantly, that t33.codes is definitively fraudulent, or that it is a crypto scam.

t33.codes is registered 2 weeks ago

Can’t verify
The available search results repeat the two-week registration claim but provide no WHOIS, registry, or other primary evidence confirming the date zeronoise.ai.

Assessment summary

  • The core claim that ChatGPT recommends a lookalike domain (t33.codes) instead of the real T3 Code site (t3.codes) is confirmed by GitHub documentation.
  • However, the characterization of the fake site as a "crypto scam" is unverifiable—sources describe it as a copied website that outranks the original in search results, not specifically as a cryptocurrency scam.
  • The "registered 2 weeks ago" claim cannot be verified.
  • The broader point about AI search enabling phishing is well-documented.

Claims reviewed

ChatGPT constantly recommends a fake crypto scam clone site (t33.codes)

Needs context
ChatGPT does recommend a lookalike domain (t33.codes) over the real t3.codes, but sources describe it as a copied website competing in search rankings, not specifically as a 'crypto scam.' T3 Code is a coding agent interface, not a cryptocurrency product. The 'crypto scam' characterization appears inaccurate.

This claim may mislead readers without added context.

t33.codes was registered 2 weeks ago

Can’t verify
No sources provide the registration date of t33.codes or confirm it was registered only 2 weeks before this claim. The GitHub issue references the problem but does not specify domain registration timing.

AI search is a new phishing vector

Supported
Multiple 2026 sources document cloned retail sites appearing in ChatGPT results, fake AI ad portals, and fraudulent websites being recommended by AI search tools, confirming this is an active threat vector.

Assessment summary

  • Parts of the statement are corroborated by multiple social reports but other parts are exaggerated or not independently verifiable.
  • Public posts show ChatGPT suggested a t33.codes link and that t33.codes is a recent domain nitter.cf xitter.cf.
  • Whether the site is a criminal 'crypto scam clone' and whether ChatGPT 'constantly' recommends it, or that this single incident proves AI search is a new persistent phishing vector, are not fully proven by the available evidence nitter.cf xitter.cf radar.cloudflare.com.

Claims reviewed

ChatGPT constantly recommends a fake crypto scam clone site (t33.codes) when users ask for the real T3 Code download link (t3.codes)

Needs context
Multiple social-media posts report ChatGPT suggesting t33.codes for T3 Code downloads, but 'constantly' overstates the available evidence and independent replication of frequency is lacking nitter.cf xitter.cf.

This claim may mislead readers without added context.

t33.codes is a fake crypto scam clone of t3.codes

Can’t verify
Community posts describe t33.codes as an almost pixel‑perfect copy of t3.codes and allege malicious intent, but I found no authoritative takedown notice, malware analysis, or formal attribution proving the site is an active crypto scam operator nitter.cf xitter.cf.

This behaviour makes AI search a new phishing vector

Can’t verify
This single reported incident shows a risk that AI outputs can surface malicious domains, but it does not by itself prove AI search is an established, systematic phishing vector across products and time nitter.cf xitter.cf.

Sources cited

16 sources

Share this result