SolutionsA single title

Readers ask their assistant, and the answer is not yours

Readers ask Claude or ChatGPT about your reporting. The subscriber who paid gets anybody's summary, the reader you never met never arrives, and blocking the crawler or licensing the archive changes neither.

The situation

What is going wrong

Readers moved into their assistants and the title did not go with them. The three usual answers, block, license or build a chatbot, each cost the reader.

The answer is not yours

Block the crawler and the reader still asks. The answer comes from somewhere else, with no reason to subscribe in it. License the archive and the AI company gets the text and the relationship.

The subscriber gets the excerpt

The reader who pays you now reads you through an assistant that was never told. The summary they get is anybody’s, and the subscription buys nothing where they read.

The new reader never arrives

A reader asks about your beat, gets a summary and never reaches your site: no sign-up, no offer. A chatbot of your own is a second index, a second login and a second copy of your rules.

What changes

When the assistant signs in to your title and reads your rules

Your title becomes a connector in Claude, ChatGPT or any assistant that takes them. The reader signs in with you, and every article is decided by the rules your paywall already runs.

EdgeKit AI Access & Answers

Your paywall’s rules decide every article

Two decisions sit in the same editor as your paywall’s rules: who may connect, and what each article returns. Full text, the standfirst and a link, or the headline alone, read off the reader’s subscriptions and the article’s section, tags, authors, type and publish date. An excerpt is the standfirst or the opening lines, never the passage that matched the question.

  • Publish a change and it reaches assistants in 60 seconds, as it reaches your paywall
  • Only published stories are indexed. A draft never reaches a model, and an unpublished story leaves the index at once
  • The index lives in your own database, beside your reader records
Decision editor · the two AI decisionsPublished

Who may connect Checked on every question

01Signed in · registered or paidAllow

Anyone else is refused.

What each article returns Checked for every article

  1. 01Section is WireHeadline and link
  2. 02Has access to MarketsSection is MarketsFull text
  3. 03Tag is Free to readFull text
  4. ElseOtherwiseExcerpt
Full text
The whole article
Excerpt
Standfirst or opening lines
Headline and link
No text
A common start: any signed-in reader may connect; full text with access to the title, otherwise an excerpt.

EdgeKit ID

Their assistant, your login, one reader record

The reader adds your title’s address to Claude, ChatGPT or any assistant that takes connectors, then signs in on a page under your name with a code sent to their email. An existing reader lands on their own account and subscriptions. A new address registers, and you meet a reader you did not have.

  • One account on your website and in the assistant
  • Each title has its own address and its own logins
  • Every reader signs in. Nothing is answered anonymously

EdgeKit Subscriptions

The subscription follows the reader into the assistant

A subscriber asks and gets the full article in the assistant they already use, because the rule reads the subscription on the same record the website reads. Access is checked on every search and every read, so a subscription that lapses loses full text on the next question. An article opened in the assistant counts on the meter, as a read on the website does.

  • Full text for the subscriber, the standfirst and a link for everyone else
  • A lapse shows on the next question, not the next sync
  • One meter, on the site and in the assistant
Meter · free articlesRead by 2 rules

EdgeKit Paywall

Every answer links to the article, where the offer is

Each result carries the article’s address on your website, so the assistant cites it and the reader clicks through to the page, where your paywall makes the offer. It is one account, not one browser session: a subscriber signs in on your website with the login they used in the assistant and reads on, and everyone else meets the offer your rules chose.

  • Search by meaning and by exact words, in the article’s own language
  • Results narrow by date, section, tag, author, type or language
  • The address is yours to publish, on a subscriber page or in a newsletter, and readers add it themselves
What the assistant receives4 results

How it goes live

Four steps, and your CMS stays the source

White-glove onboarding: we set it up with you, start to finish.

  1. 01

    Connect your archive

    EdgeKit connects to your Arc XP organization and syncs every story. We do the install, and your developers review the small integration that tells EdgeKit when a story changes.

  2. 02

    Index what is published

    Published stories are split into passages and indexed in your own database. A draft is never sent to a model, and a scheduled story is never returned.

  3. 03

    Write the two rules

    In the decision editor, like any paywall rule. A common start: any signed-in reader may connect, full text with access to the title, otherwise the standfirst and a link. Publish, and it is live in 60 seconds.

  4. 04

    Give readers the address

    Each title’s address goes on your subscriber pages or in a newsletter. Readers add it to their assistant, sign in once and start asking.

Solutions

Answer from your own archive, on your paywall's terms

Book a demo