AI Glossary for Investors: 12 Terms
Hallucination, LLM, prompt, knowledge cutoff, automation bias and more: 12 AI terms for investors, each with an investing example.
An AI glossary for investors gathers the terms worth understanding before using a chatbot near your money: hallucination, LLM, prompt, knowledge cutoff, official data, automation bias and more. Each definition is short and comes with an investing example.
Short answer: these are the AI terms that explain why a chatbot can be wrong with total confidence. Twelve short definitions, each with an investing example. For the big picture, visit the AI for investing hub.
Key takeaways
- Twelve short terms, each with an investing example.
- The golden rule: every figure with a date and a source.
- Official data wins in any disagreement with a chatbot.
See the Apple analysis in 10 seconds
What you just read, applied to Apple. A preview with real data:
- Making money? Yes
- Pricey vs its sector? Pricier
See the full Apple analysis free → Already done: nothing to type.
The terms
1. Hallucination. When a model produces something plausible but false. *Example:* it gives you Apple's P/E with no source. Check the Apple page, which dates every figure.
2. LLM (large language model). A system that generates text by predicting the next word. It does not "know" finance: it reproduces patterns. *Example:* ChatGPT is an LLM.
3. Prompt. The instruction you give it. *Example:* "Summarize the risks in this text and write "not stated" if something is missing". See examples in ChatGPT prompts for stock analysis.
4. Knowledge cutoff. The date up to which the model was trained. *Example:* if it gives you the "current" price, it may predate that cutoff.
5. Official data. A figure from a primary source: SEC EDGAR, the CNMV, the annual report. *Example:* a US company's 10-K. It wins in any disagreement.
6. Primary source. The original document, not a summary. *Example:* the annual report versus an article about it.
7. Automation bias. The tendency to trust a machine because it sounds sure. *Example:* placing an order because "the chat said so". More in risks and biases of AI stock analysis.
8. Confirmation bias. Looking only for reasons that agree with what you already think. *Example:* asking the chatbot "why will this stock go up". Ask for the opposite too.
9. Backtest. A simulation of how a strategy would have done in the past. *Example:* "it would have earned X%". It is not a promise about the future.
10. Context window. How much text the model can read at once. *Example:* a whole 10-K may not fit; work by sections.
11. Cited source. When the tool says where each figure comes from. *Example:* an analyzer that shows the date of each number, like Fintols, or N/A if it is missing.
12. Human verification. A person checks what matters. *Example:* open the source and confirm three figures before deciding. It is the conclusion of CNMV Working Paper No. 96, *Large Language Models and Stock Investing: Is the Human Factor Required?*.
Keep reading free.
Create your free account to continue reading “AI Glossary for Investors: 12 Terms”.
No card required
How to use this glossary
If a term sounds like jargon, go back to its definition and its example. Then apply the golden rule: every figure with a date and a source. To see how, read how to verify ChatGPT's financial numbers and can AI predict stock prices?. And if you want a company with official figures and N/A where data is missing, try the analyzer.
Frequently asked questions
What is an AI hallucination?
A false answer that sounds credible. In investing it can be a figure, a date or a company fact that does not exist.
What is a knowledge cutoff?
The date up to which the model received training data. It may not know anything later.
What is automation bias?
The tendency to accept what a machine says without reviewing it. You fight it by checking the source.
What is a company's official data?
What appears in its filings with the regulator: the annual report, 10-K, 10-Q or periodic reports.
Educational content; not investment advice.
Learn to analyze companies, one a week
Each week we explain a real company in simple words. Create a free account and try the analysis on a company you choose.
See it with real companies
Each of these pages shows the company's dated figures and how it compares with its sector:
Turn what you read into your first analysis.
Create a free account and run your first full analysis in a minute. No card needed.
- How it makes money
- Valuation scenarios
- What could go wrong
First full analysis free · No card
Keep learning
- Experiment: Same Questions to Several AIs (Method)
- How I Use AI Without Letting It Decide for Me
- Why NVIDIA stock is falling in November 2026
- Hardware and devices stocks
- Semiconductor stocks
- Software and cloud stocks
- All Fintols articles in English
An article from the Fintols blog, adapted for US investors. Educational content, not investment advice.