Automate Instagram DMs to potential matches
The AI automation wave just crashed into an unexpected corner of daily life - dating. Task: Developer Ben Guez used Anthropic's OpenClaw framework and Claude AI to build a script that autonomously browses Instagram, identifies potential matches, and sends direct messages. The system combines browser automation with language processing to evaluate profiles and craft personalized messages at scale. Human involvement is not explicitly detailed in the source, but the workflow implies manual setup of the automation logic and monitoring of outputs.; tools: OpenClaw, Claude. Human role: The original post does not say how a person stayed involved.; limitations: This workflow violates Instagram's terms of service, which prohibit automated access and messaging. Meta actively fights such automation, though enforcement can be inconsistent. Anthropic's usage policies prohibit deceptive activities, placing this use case in a gray area. The source does not specify technical constraints or costs, but the risk of account suspension is high..
Tools
Published on Agent Work Atlas:
Overview
The AI automation wave just crashed into an unexpected corner of daily life - dating.
Workflow
Developer Ben Guez used Anthropic's OpenClaw framework and Claude AI to build a script that autonomously browses Instagram, identifies potential matches, and sends direct messages. The system combines browser automation with language processing to evaluate profiles and craft personalized messages at scale. Human involvement is not explicitly detailed in the source, but the workflow implies manual setup of the automation logic and monitoring of outputs.
Reported result
The source reports that the script filled Guez's Instagram message requests with what he describes as "a bunch of potential international wives." No measured outcomes regarding actual relationships, response rates, or account longevity are provided. The success is anecdotal and limited to the volume of incoming messages generated by the automation.
Human role
The original post does not say how a person stayed involved.
Limitations
This workflow violates Instagram's terms of service, which prohibit automated access and messaging. Meta actively fights such automation, though enforcement can be inconsistent. Anthropic's usage policies prohibit deceptive activities, placing this use case in a gray area. The source does not specify technical constraints or costs, but the risk of account suspension is high.
Try it
I want to experiment with AI-driven social automation using Anthropic's OpenClaw and Claude. My goal is to [replaceable: define specific objective, e.g., 'test profile engagement strategies']. I have access to [replaceable: list of target platforms, e.g., 'Instagram test accounts']. Please guide me through setting up a script that [replaceable: specific action, e.g., 'analyzes public profiles and drafts messages']. Steps: 1. Review the ethical and legal implications of automating interactions on [replaceable: platform name]. 2. Draft a Python script using OpenClaw for browser automation and Claude for message generation. 3. Create a dry-run mode that logs actions without sending messages. 4. Provide a sample output for [replaceable: number] profiles. Deliverables: A script file, a log of the dry run, and a risk assessment report. Quality Check: Ensure the script does not violate platform ToS in the dry run. Error Handling: If the platform blocks the automation, stop and report. Safeguards: Do not send any real messages or perform any actions on live accounts without my explicit confirmation after reviewing the dry run. Ask clarifying questions if the target platform or objective is unclear.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.