Corrections

This page is a public log of every correction we have made to AI Trade Lab content. Anywhere we cite specific numbers, timestamps or official sources, we run a quarterly spot-check, plus immediate re-verification whenever a reader emails us about a possible error. Each entry shows the old text, the new text, the date and the reason — so you can see exactly what we changed and why.

1. How corrections work

Quarterly spot-check: In the first month of every quarter we audit a sample of the previous quarter's articles, focusing on paragraphs that cite numbers, dates or official policies.
Reader-triggered review: Send corrections to privacy@bnhous.com. We acknowledge and complete review within seven working days.
Official policy monitoring: When Binance publishes a fee or policy change, we re-check related articles.
AI tool version monitoring: When the AI tools we review (ChatGPT / Claude / Gemini and others) ship a major version, we re-run the relevant evaluation paragraphs.
Correction notice: A visible Correction Notice is added to the top of the affected article. Material fact corrections that may affect reader judgement are highlighted on this page for at least 30 days.
Permanent log: Past corrections are never removed from this page.

2. Correction log

2026-05-20 · Binance Grid Bot return-rate range corrected

Affected article: Binance AI features guide and the Grid Bot sections, plus prompt-library strategy reviews.

Old text: Binance Grid Bot historical backtests achieve 15%–30% annualised returns.

New text: Binance Grid Bot annualised returns are highly volatility-dependent: in our hands-on tests with BTC/USDT and ETH/USDT through H2 2024, results ranged from 5% to 25% annualised — high-volatility, range-bound pairs with a narrow grid could hit 20%+, while trending markets or low-volatility narrow ranges often came in near zero or even drawdown. Backtests do not guarantee future performance; evaluate against current volatility yourself before deploying.

Reason: The old wording ("15%–30%") came from Binance marketing copy and a few bloggers copy-pasting it, without distinguishing market regime or volatility tier. A reader emailed that their actual Grid Bot run during a 2024 downtrend went underwater, not matching our wording. We re-ran real historical backtests across six pairs, found the lower bound should be 5% and that 30% is not a stable upper bound. The new copy carries the measured range plus an explicit "backtests ≠ future performance" guard.

3. How to submit a correction

Spotted something inaccurate? Email privacy@bnhous.com and include:

The URL of the article in question
The paragraph or claim you think is inaccurate
Your suggested wording, ideally with an official source link or reproducible test conditions

Confirmed corrections are applied immediately and logged on this page. If you agree to be named, we credit you in the entry.

4. Related pages

Editorial Standards — full source hierarchy, review flow, AI-content labelling, rating methodology and corrections process
About — team, independence and Affiliate relationships
Affiliate Disclosure — revenue model and editorial-independence boundaries

Crypto plus AI trading tools is a YMYL plus rapidly-shifting technical area — specific numbers, dates, policies or model performance can change with any upstream adjustment or model release. Re-verify on the official page or the tool's latest version before acting. All AI Trade Lab content is informational and is not investment, tax or legal advice.