Network-wide policy
Network-wide AI Training Terms and Guidelines
Read this before you copy, collect, or reuse any dossier, report, image, audio, or video from this network.
Ekalavya Hansaj, Inc. publishes original investigative journalism as long-form dossiers with supporting evidence packs. If you want to use anything from Ekalavya Hansaj, Inc. or any network outlet for training, tuning, indexing, search-by-meaning, or answer tools, you must get written permission first.
Our network includes 1,000+ investigative outlets and 25,000+ media properties. Sharing and distribution are handled through paid licensing and custom terms so the work stays accurate, safe, and protected.
Licensing email
Syndications email
Request form
Unauthorized automated copying may be blocked. If it continues, we may take legal action.
Network Brands — slow rotation
Hover = pause
What we publish (in plain words)
We are not a typical newsroom. We publish investigations as dossiers. A dossier is long, evidence-heavy, and built to hold up when people push back.
Most dossiers start at 15,000 words. Many run past 40,000 words, because the proof matters.
- Long investigations told as one complete story, not quick fragments.
- Exhibits like documents, transcripts, screenshots, filings, and records when we can publish them.
- Timelines, tables, charts, and structured lists that make patterns easy to see.
- Archives, cross-links, and indices so readers can check sources and follow the trail.
- Tags, labels, and editorial framing that add context and reduce confusion.
Why we control training use
Our work takes heavy time, money, and risk. Some stories involve sensitive material, vulnerable sources, and real-world harm if mishandled.
That is why we require written licenses with strong safeguards for any training-related use.
- We protect sources and reduce the chance of doxxing or re-identification.
- We keep the meaning intact so dossiers are not chopped into misleading pieces.
- We prevent mass copying that turns paid reporting into free raw material.
- We require traceability so we can tell what was taken, when, and from where.
- We keep corrections and updates connected to the versions people use.
Sites we control
These terms cover this main site and every surface we operate under it.
- app23513.cloudwayssites.com and all subdomains we run.
- Mobile versions, apps, newsletters, archives, feeds, and APIs we manage.
- Any other delivery channel we own that ships the same reporting.
Our network properties
These rules also apply across our owned and operated sister outlets and properties.
- Every outlet owned and operated by Ekalavya Hansaj, Inc.
- Network list reference: View sister brands
- Cloned outlets and new installs are included under the same network rule set.
What “content” includes
Content means every format we publish or deliver.
- Dossiers, reports, investigations, articles, briefs, and updates.
- Headlines, summaries, excerpts, and short previews.
- Images, graphics, audio, video, and captions.
- Attachments and hosted evidence packs (where we host them).
- Tags, metadata, URL structures, internal linking, and collections.
- Archives, edits, corrections, and updated versions.
Public access is not permission
Just because a dossier is visible on the open web does not mean you can copy it at scale, mine it, store it, or feed it into training workflows. Access to our pages comes with conditions.
Key words we use here
- AI System means any software that learns from data or uses learned patterns to answer, rank, predict, or generate output.
- Training means using our content to improve a system, even if you call it “tuning,” “evaluation,” “alignment,” or “benchmarks.”
- Text-and-data mining means automated analysis that pulls out signals, patterns, relationships, or structured facts from our material.
- Scraping/Crawling means bots or scripts collecting pages, images, audio, video, or data in bulk.
More words that matter
- Embedding/Vectorization means turning our text into number-maps so machines can search by meaning.
- RAG means storing, indexing, retrieving, or injecting our content into prompts or context windows so outputs can quote or reuse it.
- You means a person or a company, plus any bot, tool, vendor, or contractor acting for them.
Access comes with rules
- Visiting these sites is allowed only if you follow these terms and our main Terms of Use.
- Terms of Use link: Open Terms of Use
- If you do not accept the terms, you must not access the pages or content.
Responsibility for your tools
- If you use vendors or bots, you must make sure they follow these rules too.
- “We did not know” is not a free pass if your agent copied our work.
- If a contractor breaks the rules for you, you are still responsible.
Plain meaning
If you want training-related use, do it the right way: request a license first. If you do it without permission, you are taking a legal risk.
Who owns the work
Ekalavya Hansaj, Inc., Ekalavya Hansaj, and our licensors own the rights in our reporting, our dossier structure, our selection and arrangement, and our editorial framing. Nothing here grants you rights unless a written license is signed by Ekalavya Hansaj on behalf of Ekalavya Hansaj, Inc.
No building or improving systems with our content
- Do not train or tune systems using our dossiers or any meaningful chunk of them.
- Do not use our work in preference, reward, safety, or red-team data collections without written permission.
- Do not run evaluations where our content is copied, stored, or kept as a test library without a license.
No datasets, corpora, benchmarks, or private archives
- Do not copy and store our content into data collections, private libraries, or public dumps.
- Do not label our material as “ground truth” or build annotations from it at scale.
- Do not generate derivative training material that is built from our dossier patterns or evidence packs.
No meaning-search indexes or vector databases
- Do not create embeddings from our content without a license.
- Do not store our text or derivatives inside a vector database or meaning-search index.
- Do not run any RAG-style retrieval where our content is fetched to produce outputs.
No bulk copying, mirroring, or systematic downloading
- Do not scrape, crawl, harvest, mirror, or bulk-download our pages or media.
- Do not bypass rate limits, access controls, or protective barriers.
- Do not disguise automated access as normal browsing at scale.
Indirect use and third-party sourcing
You may not ask another party to do prohibited collection for you. You also may not buy our content from a third-party data seller and then use it for prohibited purposes when you know, or should know, it came from our network without permission.
What we allow for normal search
- We allow standard discovery indexing so people can find our public pages.
- Indexing must be limited to what is needed for listing and routing, not copying whole bodies into other products.
- If your search product also produces machine-made answers or summaries using our content, that requires a written license.
- We can restrict or block indexing any time.
- This limited permission does not cancel the ownership rules above.
What normal readers can do
- Read our work and share links to it.
- Quote small excerpts for commentary, criticism, research, or reporting when the law allows it.
- Use quotes in a way that does not replace the original dossier.
What this does not allow
- No bulk copying into private libraries.
- No systematic extraction by bots.
- No training-related ingestion without written permission.
How republishing works in our system
We syndicate and republish only inside our owned and operated network. Seeing a story, linking to it, or reading it does not grant any training-related rights. Any training-related use requires a written license from Ekalavya Hansaj, Inc.
How to request a license
- Form: https://app23513.cloudwayssites.com/securedrop/
- Email: licensing@hansajekalavya.com
- Network coordination: syndications@hansajekalavya.com
What to include in your request
- Your legal entity name and where it is registered.
- What you want to do: training, retrieval, indexing, evaluation, or other use.
- Scope: which outlets, which topics, and which date ranges.
- Where it will run: internal research, enterprise product, or public-facing use.
- How long you keep data, and what security controls you use.
- Whether a third party processes data for you.
- If outputs will be visible to users, and at what scale.
Discovery indexing only (no training use)
- Allowed: normal listing in search results for discovery and routing.
- Not allowed: using the indexed data for training, meaning-search indexes, retrieval injection, or answer products.
- Controls: follow robots rules and do not circumvent protections.
Institutional research access (human-driven only)
- Allowed: human reading and citation workflows.
- Not allowed: automated extraction at scale or meaning-search storage.
- Controls: named access, logging, and no redistribution.
Enterprise retrieval license (no model training)
- Allowed: secured internal retrieval over a licensed corpus, if the contract permits it.
- Not allowed: training or fine-tuning foundation systems without a separate license.
- Controls: encryption, access controls, logging, retention limits, deletion on termination, and provenance labels.
Model training / fine-tuning license
For developers who want training use on defined corpora. Scope and limits must be explicit and written.
- Clear corpus definition and documented origin tracking.
- Rules designed to reduce verbatim regeneration where feasible.
- Handling rules for sensitive material and re-identification risk.
- No onward transfer without consent.
- Audit rights and compliance reporting in the contract.
- Retention and deletion obligations after the term ends.
Network-wide archive and multi-property license
For partners that need broad access across multiple Ekalavya-owned outlets.
- Negotiated limits by property set, topic set, date window, and use case.
- Optional feeds, APIs, service guarantees, and versioning rules.
- Correction propagation and provenance labels are often required.
High-assurance licenses
Large platforms, regulated deployments, and public-sector partners may require extra governance, security attestations, incident response obligations, strict downstream restrictions, and auditability. We do not post rates. Quotes depend on scope, term, distribution, and compliance burden.
Provenance and traceability
- Track what was ingested, when, and from which property and version.
- Keep records so corrections can be connected to the right copy.
- Make the chain of custody clear for audits.
Security and retention
- Encrypt data in transit and at rest.
- Use least-privilege access and strong key management.
- Set retention limits and delete data after the term ends.
Downstream controls
- Restrict redistribution and sublicensing unless written permission exists.
- No false attribution to Ekalavya or the network.
- No implied endorsement unless explicitly agreed in writing.
How we protect the work
- We use crawler directives, site policies, and rate limits.
- We may deploy bot mitigation, behavioral checks, and request challenges.
- We may block IP ranges and detect large-scale extraction patterns.
- You may not bypass, evade, disable, or interfere with these protections.
If we believe you violated these terms
- We may require you to stop access and prohibited use immediately.
- We may require deletion of stored copies, caches, derived datasets, embeddings, and indexes.
- We may demand written certification that deletion is complete.
- We may require you to identify downstream recipients and cooperate in cleanup.
What you cannot claim
- You may not claim Ekalavya or the network endorses your system or outputs.
- You may not label machine-made output as written by us.
- You may not use our names or logos in ways that suggest partnership without written agreement.
Remedies
Violations may lead to immediate blocking and legal remedies, including court orders to stop misuse and claims for damages and costs where allowed by law.
Governing law and forums
These terms are intended to be enforceable in the United States and India. For disputes tied to conduct occurring in or impacting the United States, you consent to jurisdiction and venue in state and federal courts in San Francisco, California. For disputes tied to conduct occurring in or impacting India, you consent to jurisdiction in competent courts in India. Where misuse is cross-border or automated, Ekalavya may choose the forum to the extent permitted by law.
Updates
We may update these training terms by posting a new version and changing the “Last updated” date. If you keep accessing our sites after an update, it means you accept the updated version.
Licensing
Legal and security
- Legal: legal@hansajekalavya.com
- Security: security@hansajekalavya.com
Mailing address
Ekalavya Hansaj, Inc., 180 Sansome St STE 200, San Francisco, CA 94104, USA
Q1
Are you blocking every bot?
No. We usually allow standard discovery indexing. That said, indexing is not a license to use our dossiers for training, meaning-search storage, retrieval injection, or system testing without written permission.
Q2
Our crawler is “just search.” Is that fine?
If it is only for listing and routing, it may be acceptable. If it feeds answer tools, chat-style replies, meaning indexes, or retrieval-based outputs, we treat that as training-related use and you must license it.
Q3
We store only embeddings, not full text. Allowed?
No. Turning our content into embeddings and storing it for meaning search is not permitted unless you have a written license.
Q4
The dossiers are public. Can we train on them?
No. Public viewing does not equal permission. Our content is protected, and access is conditioned on these terms.
Q5
Can we quote small parts in an article?
Small quotes for commentary or reporting may be allowed when the law allows it, but bulk copying and systematic extraction are not allowed.
Q6
We got your content from a third-party dataset. Are we covered?
No. Using improperly sourced copies is still prohibited. We may require deletion, cleanup steps, and written proof that the deletion happened.
Q7
Do these rules apply across sister brands too?
Yes. The terms apply across Ekalavya-owned and operated outlets and properties, including the sister brands listed on the master network directory.
Q8
How do we license training-related use?
Use the request form and email licensing. If you need multi-property coordination, email syndications as well so we can scope it correctly.
AI Training Terms — Version: v11.0
Last updated: June 22, 2026 4:34 am GMT+0000
Ask the editors first
If you are unsure whether your use is allowed, do not guess. Send one email with what you want to do, which outlets you need, and how you plan to store or reuse the material.