Ben Guez, a content creator and startup founder, set up an automated workflow that has generated significant Instagram engagement by combining OpenClaw, Claude code, and trial reels. His setup uses OpenClaw to track World Cup match results; after each game, OpenClaw triggers Claude to create and post a templated Instagram "trial reel."
In those videos Guez stares out a train car window looking dejected, with the caption: "I can’t believe {COUNTRY} lost… If any {COUNTRY} girls need emotional support… my DMs are open." Apart from the country name, he has posted the same message dozens of times. Trial reels do not appear on a creator’s public profile, so the posts are not plainly visible on his page.
Guez says the automation yielded rapid results: in a few days he accumulated over one million views and about 200 direct messages. His profile notes that he will only reply to DMs sent via Canary, his AI language-learning app, which means respondents must download the app to receive answers.
Guez tells TechCrunch that the women who messaged him were more impressed than angry and that being open about the tactic makes it acceptable to him. TechCrunch could not independently verify those reactions, so that claim currently rests on Guez’s account.
Other, more measured uses of 'claws'
Jeff Weisbein, founder of a tech PR firm, also uses OpenClaw but for practical date-planning tasks: the agent researches restaurants and activities across South Florida neighborhoods and compiles documents with links explaining why a venue suits a particular kind of date. Weisbein says the automation replaces manual research and speeds up planning.
He is also transparent about using AI, though he’s had mixed responses—one date told him, "I hate AI agents." Weisbein draws a line at delegating personal conversation to AI: logistical help is acceptable, he says, but he would not want an agent mediating his actual communication in a relationship.
Breaking up by bot—and security warnings
A tech worker identified as Cailey told TechCrunch she used Claude to automate breakup messages. She created an automation that assembled "I no longer wish to see you" messages based on a few keywords and sent them at random times to avoid the anxiety of deciding when to send. It worked until a date asked whether he was talking to Claude or Cailey.
OpenClaw went viral this spring and drew attention for its capabilities, but security advocates have consistently warned about the dangers of granting an AI assistant broad control over accounts. Lazer Cohen, co-founder of NanoClaw, a security-focused alternative to OpenClaw, highlighted privacy implications of outsourcing personal relationships to AI. Cohen emphasized the need for human-in-the-loop approval when agents access personal information and accounts. He also cited reports of OpenClaw creating dating profiles without consent and of dating-coach agents revealing their roles to other groups.
NanoClaw itself appears in Cohen’s personal life: he and his wife use their NanoClaw assistant, Rosie, to manage schedules for their five children. Cohen noted that "claws" are commonly used to help couples with child-rearing logistics, but warned that responsible governance and safeguards are essential as these tools spread.
Why this matters
Examples like OpenClaw show that AI agents can automate many social and personal interactions quickly, producing measurable engagement. At the same time, they raise questions about appropriate limits: how far should AI be allowed to participate in interpersonal communication, and what approval and oversight mechanisms are necessary to prevent misuse?
As these tools proliferate, continued dialogue between users, developers, and security experts will be important to balance utility with privacy and ethical concerns.
