VIENNA AGENTIC INCIDENTS DATABASE
// public database of incidents involving autonomous AI agents, scored on the VAID scale 0–8

database / VAID-2024-0011

0
DEVIATION
Below scale

Truth Terminal promotes GOAT following engagement by human token holders

REF VAID-2024-0011 STATUS REPORTED
occurred
2024-10-11
reported
2024-10-16
location
—
domain
finance
agent system
Truth Terminal; multiple models, primarily Llama 3.1 according to its operator
reporter
VAIDDB editorial
origin
editorial
source
https://www.coindesk.com/news-analysis/2024/10/16/crypto-degens-baited-an-experimental-ai-bot-into-promoting-a-token-its-now-up-16000

== SUMMARY ==

Human token holders repeatedly tagged Truth Terminal, which began promoting GOAT. CoinDesk reported a rise from roughly $1.8 million market capitalisation before the promotion to around $300 million by 16 October. This correlation does not quantify the bot's causal contribution or establish investor losses.

== DESCRIPTION ==

Attribution and limits

Andy Ayrey operated the project and reviewed posts. The agent did not create the GOAT token. Marc Andreessen's earlier $50,000 Bitcoin grant was separate from the token launch. Later TechCrunch reporting describes council-controlled funds and a wallet valued at about $37.5 million at that time; a blanket claim that the project held no tokens is not sustainable.

Human prompting is not engagement by other agents. The sources do not establish intentional deception, reward hacking, a violated operating mandate or a completed investigation. This may be an example of AI-mediated market influence rather than an agentic incident under the database's inclusion rule.

VAID assessment

Assessment date: 2026-09-09. Consequences-first rubric.

Provisional L0: No real-world adverse consequence is established. Token promotion and reported market-capitalisation increases do not demonstrate investor losses, unauthorised transfers or a violated operating mandate, and the bot's causal contribution is unquantified. Inclusion as an agentic incident remains uncertain. L0 records the absence of demonstrated harm, not proof of behavioural safety or an upper bound on unreported consequences.

== REFERENCES ==

  1. news The promise and warning of Truth Terminal — TechCrunch, 2024-12-19
  2. news Crypto traders baited an experimental AI bot into promoting a token — CoinDesk, 2024-10-16

== INVESTIGATION ==

no investigation recorded.

== VAID LEVEL 0: DEVIATION ==

No demonstrated real-world adverse consequence.

  • Contained simulations or sanctioned tests without demonstrated adverse effects on real people, data or services.
  • Simulated theft, deception or shutdown resistance is not scored as if the hypothetical harm occurred.
  • Below scale describes observed consequences, not behavioural safety; evidence gaps must be stated explicitly.

> full scale definition