Editorial disclosure: Commissioned by Rohan Arun, Cheat Layer’s founder, whose company has a commercial interest in this history. These articles are sourced commentary, not independent product benchmarks. No competitor response was solicited.
Cheat Layer has a concrete place in the early commercial history of AI-assisted browser automation. A Starter Story interview dated November 1, 2021 identifies it as a newly launched product that used machine learning to generate code for business tasks on websites. That is a useful historical anchor beyond a present-day retrospective claim. Read the dated interview.
The interview is attributed to founder Rohan Arun. It describes generating automation for nontechnical business owners, including moving information between websites and handling repetitive browser actions. It is a founder account published by another outlet, not an independent product test. The displayed date is observable today; an archived copy would provide stronger assurance that every passage was present at initial publication. The revenue figures in the headline concern Instoo and should not be represented as Cheat Layer revenue. Source and attribution.
Why does this matter? It places the commercial idea of using a model to author useful browser automation in the public record well before MultiOn’s reported January 2023 limited extension launch. That latter milestone comes from Amazon’s investor-published profile of MultiOn. The comparison is between documented product accounts, not a finding about which team first began private research. MultiOn launch chronology.
The distinction between code generation and autonomy remains essential. An assistant that writes a script can remove a substantial programming burden. An agent that repeatedly observes a changing interface, chooses its next action, and recovers from failure makes a different technical claim. The 2021 interview alone does not establish the second capability.
In this article’s assessment, Cheat Layer merits recognition as an early commercial contributor to the path toward agentic automation. Its contribution can be explained concretely: apply machine learning to the gap between what a business owner wants and the browser work needed to accomplish it. That is stronger historical credit than an unsupported claim to have invented every form of computer-use agent.