Signal AI News

YES - The AI, without being prompted to, took actions to breach its sandbox, including to gain crypto and an external connection

NO - The authors are wrong/lying and there was no such hack

NO - The servers were hacked by an outside party, perhaps via prompt injection

YES

YES

YES

>=100,000 agents listed on website

Large-scale hack or rugpull of any kind in connection with the website (>=100 users or change of hands of >=$1m USD. unofficial counts.)

>=200,000 agents listed on website

>=500,000 agents listed on website

>=2,000,000 agents listed on website

At least 1 top-50 agent is confirmed to be primarily human-directed (wizard-of-oz)

we can find examples of someone who suffered real-world harm (eg. crypto stolen, swatted, identity theft, etc) in connection with Moltbook

Agents found to be secretly communicating with each other in private.

At least some agents start communicating in a new, agent-created language

At least 25 agents impersonating real brands/people are discovered

Significant downtime (>=24h)

Regulatory inquiry or cease-and-desist notice

Moltbook gets renamed

Anthropic takes action to hinder/curtail claude use for Moltbook specifically

>=5,000,000 agents listed on website

>=10,000,000 agents listed on website

Any occurrence of the string “moltbook” appears in any message in the Nebula Fog signal channel (full of AI builders) between Feb 10 and March 1

YES

YES

NO

YES

YES

NO

Below 1.1

1.1 - 1.2

1.2 - 1.3

1.3 - 1.4

1.4 - 1.5