Fact-Checking Policy

How BlockAI News verifies claims, sources, and on-chain data before publication.

Last updated: May 9, 2026

BlockAI News covers two of the most misinformation-prone domains on the internet — crypto and artificial intelligence. Inflated claims, fake announcements, and unverified on-chain transfers are routine. This policy describes the standards we apply to every article published on blockainews.com.

Our Standard

Every claim of fact in a BlockAI News article must be either:

  1. Directly verifiable from a primary source (an official announcement, regulatory filing, GitHub commit, on-chain transaction, or first-party statement on X/Twitter), or
  2. Confirmed by at least two independent reliable sources when no primary source exists.

We do not publish single-sourced claims about market-moving events unless the source is an authoritative primary source — such as a regulator, company filing, official announcement, or verifiable on-chain record. If we cannot meet this bar, we delay or drop the story.

What We Verify

Before publication, our editors check:

  • Quotes — verified against the original recording, transcript, or post. We do not paraphrase quotes inside quotation marks.
  • On-chain data — wallet addresses, transaction hashes, and token transfers are verified by querying the relevant blockchain explorer (Etherscan, Solscan, Dune, Arkham, etc.) before we cite a number.
  • Funding rounds and valuations — confirmed via the lead investor's public announcement, the company's official channels, or SEC filings. We do not republish leaked numbers from group chats.
  • Technical claims about AI models — verified against the official model card, paper, or repository. Benchmark numbers are taken from the original publication, not aggregator screenshots.
  • Regulatory actions — confirmed against the regulator's official document (SEC, CFTC, OCC, ESMA, FCA, etc.), not a third-party summary.
  • X posts — verified directly against the original post at publication time. If a post is deleted or unavailable, we either remove it, label it accordingly, or rely on archived evidence where appropriate.

Primary Sources Whenever Available

Every News article on BlockAI News cites primary sources whenever available. For market-moving or disputed claims, we require either direct primary-source verification or confirmation from at least two independent reliable sources. We do not treat reporting from competing crypto outlets as primary; we link to them only as context, never as the sole basis for a claim.

The full sourcing list appears at the bottom of every News, Editor's Pick, and Learn article under a clearly labeled Sources heading.

We do not treat deleted or inaccessible posts as verified sources unless we have reliable archived evidence and clearly disclose the limitation.

Use of AI in Verification

We use AI tools (Claude, ChatGPT, web-search agents) to assist with research, drafting, and source discovery. We do not rely on AI to verify facts. Every claim attributed to a source is checked against that source by a human editor before publication. For our full disclosure on AI usage, see our AI Use Policy.

No Fabricated Ratings

BlockAI News does not publish star ratings, "Editor Scores," or aggregate review numbers unless we have a documented, public methodology behind them. Where no methodology exists, we omit the score entirely.

Limits of What We Verify

Some categories of information cannot be independently verified at publication time. We label these clearly:

  • Unaudited financials — token supplies, treasury holdings, and revenue figures provided by projects without audited statements are attributed to the source and noted as unverified.
  • Anonymous founders — we do not pretend to know the real identity of pseudonymous teams.
  • Forward-looking statements — roadmap promises and TGE timelines are reported as claims, not facts.

When We Get It Wrong

Despite this process, errors happen. When they do, we follow the procedure described in our Corrections Policy: corrections are timestamped and explained clearly. When appropriate, we describe the original error without preserving harmful, defamatory, or unsafe material. Material corrections trigger an editor's note at the top of the article.

To report a factual error, email corrections@blockainews.com or jason@blockainews.com.