About Aposema
Aposema works for publishers who earn from ads or subscriptions. It measures how much of their work AI products already take, then puts their terms in front of each of those products and enforces them.
Why it exists
AI answer products read publishers' articles and research, then answer the reader on their own page. The visit that would have shown an ad or sold a subscription never happens. Microsoft's own figures, unsealed in the New York Times' case against Microsoft and OpenAI, put click-through from Copilot's answers to the Times and the Daily News 83–93% below Bing search (Digiday, September 2026).
Most publishers rely on robots.txt, which asks crawlers to stay away. In December 2025 the federal court hearing Ziff Davis v. OpenAI found that robots.txt files do not "effectively control" access to content, which is the test for protection under the DMCA's anti-circumvention rule. Aposema attaches each publisher's copyright notice and licensing terms to every article, licenses the AI products that pay, and deceives, poisons or traps the ones that ignore the warning. Every unlicensed read goes on the record for a licensing or legal claim.
What's behind it
Aposema's technology is patent pending. The legal argument is set out in its whitepaper, Compound Statutory Liability Entrapment in Inference-Time AI Pipelines (SSRN, 2026), which analyzes the DMCA liability AI answer products face when their pipelines strip copyright information from the pages they fetch.
Aposema takes part in the IETF AIPREF working group, which is writing the shared vocabulary for AI-use preferences, and works with a collective-licensing partner that negotiates deals for many publishers at once.
The name
Aposematism is the warning coloration of animals such as the poison dart frog: bright colors that tell a predator an attack will cost more than it gains. Aposema gives a publisher's pages the same kind of warning, written as terms AI products can read.
Questions? Email team@aposema.com.