---
language: en
title: "Artificial general intelligence"
description: "AGI refers to artificial intelligence with broad capabilities across different tasks. Learn about the competing criteria and how to assess claims of generality."
canonical: "https://tropk.ai/en/resources/glossary/artificial-general-intelligence"
last_modified: "2026-10-12T16:13:28.268Z"
---

# Artificial general intelligence

Reviewed on 2026-10-06.

## What it means

AGI stands for artificial general intelligence. The concept describes systems with broad capabilities for performing different intellectual tasks, with performance comparable to humans according to the criterion used. There is no single definition shared by all researchers. Assessing a claim of AGI requires specifying the range of tasks and the level of performance considered.

## Why it matters for marketing

In marketing, the term appears in vendor presentations and forecasts about work. A promise of generality does not indicate whether a system performs well at research, editorial review, or campaign analysis. Teams can turn this claim into verifiable questions about tasks, conditions, and failure modes, rather than basing budgets or operational responsibilities solely on the label used in a demo.

## Example in practice

An agency receives a tool presented as AGI for planning campaigns. In the hypothetical demo, it writes fluent ads but fails to identify an inconsistency between the budget and channel allocation. The agency decides to assess writing, calculation, and brief interpretation separately. The claimed breadth needs to be examined alongside performance in each domain, not inferred from the fluency of a piece of text.

## How to apply or verify

When assessing a claim, ask for the definition used, the task set, the human benchmark, and the results for each task. Check whether the test measures situations close to the intended use and whether it requires tools or assistance. The Levels of AGI proposal separates generality from performance, helping people discuss capabilities without treating a single demo as certification.

## Important distinctions

AGI is not synonymous with [generative AI](/en/resources/glossary/generative-ai), autonomy, or consciousness. Producing images and text describes a function; operating without intervention describes a condition of use. In the classification by Morris and coauthors, consciousness is not a requirement, and autonomy is analyzed separately. This is a classification proposal, not a universal consensus, a timeline prediction, or proof that a product has achieved AGI.

## Sources and references

- [Levels of AGI: Operationalizing Progress on the Path to AGI](<https://arxiv.org/html/2311.02462v2>) — Meredith Ringel Morris e coautores, Google DeepMind / arXiv

## Continue a leitura

- [Artificial intelligence](</en/resources/glossary/artificial-intelligence>)
- [LLM](</en/resources/glossary/llm>)
- [Frontier model](</en/resources/glossary/frontier-model>)

## About the author

[Vinicius Pacheco](</en/team/vinicius-pacheco>)

Responsible for Tropk’s technical content, focusing on crawling, indexing, structured data, and agent access to content.

---

[All pages in Markdown](<https://tropk.ai/sitemap.md>)