Automated Access Policy

1. Publications are open

Search engines and AI systems may crawl, index, retrieve, summarise, quote and mine our publications, including to train AI models, and cite them with attribution to AlphaCitation. Our publications are the Insights publications and research papers (/insights), with their PDFs and evidence packages (/research), the Methodology page, the Label Ledger page and the other public pages listed in our sitemap, except the two sample scorecards under /scorecard, which are part of the product. Our robots.txt allows every crawler to fetch them, and also names thirteen AI user agents individually, from OpenAI, Anthropic, Google, Perplexity, Amazon, Apple and Common Crawl. This permission covers our own material; third-party material quoted in a publication remains its owners'.

Where a publication says so, the original material for which Akunudo LLC-FZ holds the rights is licensed under CC BY 4.0, or under another licence the publication names (the approved scripts in Paper No. 01's evidence package are under MIT). Quoted third-party text, third-party data fields, API/provider-restricted material and third-party marks are excluded from that grant, as each publication states.

2. The product is not open to automated access

The product means the scorecards and reports (/scorecard), customer accounts (/account), the administration area (/admin), every programming interface (/api), the sign-in and purchase pages (/login, /auth, /thanks), the free scan, and the report assistant and its live questions to the AI engines. Automated access to any of these needs our written permission, whatever the tool, including browser automation. Three uses need none:

• a search engine indexing and linking an individual free scorecard page that robots.txt allows;
• an AI agent acting for one person, at that person's request and pace, that identifies itself as section 4 describes; and
• verifying an individual Evidence Passport through its public verification page or endpoint.

3. What is prohibited without written permission

Without our written permission, you may not:

• scrape, copy or harvest product pages, reports or programming-interface responses;
• enumerate scorecard, report or other addresses, or probe for ones that are not linked;
• bulk-download scorecards, reports or any other product output;
• use product outputs to train or fine-tune an AI model, or to build a dataset, index or service that substitutes for the product;
• reverse-engineer, decompile or extract our software, prompts or classifiers;
• run autonomous agents against the product without disclosing that they are automated and who operates them; or
• circumvent rate limits, bot verification, access controls or the robots.txt rules.

Independent evaluation is permitted. You may use product outputs you obtain lawfully to evaluate the accuracy, consistency, sourcing and behaviour of AI systems, including ours, and publish your methods and findings, including findings unfavourable to us. This does not allow training a model on the outputs, building a dataset or index that substitutes for the product, republishing outputs in full beyond short excerpts that evidence a finding, or circumventing rate limits, bot verification or access controls.

As rightholder, we expressly reserve the use of the product's content for text and data mining under Article 4(3) of Directive (EU) 2019/790. The reservation is stated in machine-readable form in the tdm-reservation response header and in /.well-known/tdmrep.json. It does not apply to the publications described in section 1, and it does not affect the exception for scientific research in Article 3 of that Directive.

4. How automated agents should identify themselves and behave

Automated agents should:

• send a user agent that names the agent and its operator, with a contact address or a page describing it;
• sign requests with Web Bot Auth where their platform supports it, so they can be verified;
• follow robots.txt; and
• keep request rates modest.

We rate-limit the product and verify requests to the endpoints that run paid AI queries or start a purchase; requests that fail verification or exceed a limit may be challenged or refused.

5. Asking for permission

To ask for automated access to the product, for research, an integration or anything else, write to ozue@alphacitation.com. Tell us who you are, what you want to access, how often, and why. Security reports go to the same address; see /.well-known/security.txt. Other questions about this policy: analyst@alphacitation.com.

6. How this policy relates to our Terms

This policy forms part of our Terms of Service. If you have a subscription, section 6 of the Terms sets out what you may do with your own reports: they are licensed to you for your internal business use.