# Center for Human-Compatible Artificial Intelligence

> AI safety research center

**Wikidata**: [Q85751153](https://www.wikidata.org/wiki/Q85751153)  
**Wikipedia**: [English](https://en.wikipedia.org/wiki/Center_for_Human-Compatible_Artificial_Intelligence)  
**Source**: https://4ort.xyz/entity/center-for-human-compatible-artificial-intelligence

## Summary
The Center for Human-Compatible Artificial Intelligence (CHAI) is an AI safety research center and research institute based in Berkeley, United States. Founded in 2016, the organization is affiliated with the University of California, Berkeley, and focuses specifically on the field of AI alignment.

## Key Facts
- **Also Known As:** CHAI
- **Inception:** 2016
- **Location:** Berkeley, United States
- **Affiliation:** University of California, Berkeley
- **Primary Field of Work:** AI alignment
- **Organization Type:** Research institute, research center
- **Website:** http://www.humancompatible.ai/ (English)
- **Twitter Handle:** @chai_berkeley (User ID: 1058055843466244096)
- **Social Media Followers:** 1,229 (as of January 4, 2021)
- **Google Knowledge Graph ID:** /g/11j1c7dyr4
- **Wikipedia Coverage:** Available in 2 languages (English, French)

## FAQs
### What is the primary purpose of the Center for Human-Compatible Artificial Intelligence?
The center functions as a research institute dedicated to generating new knowledge through systematic investigation. Its specific field of work is AI alignment, distinguishing it from organizations that focus on policy or commercial applications.

### Where is the center located and who is it affiliated with?
The organization is located in Berkeley, United States, and maintains an affiliation with the University of California, Berkeley.

### How is this organization different from a standard university department?
As a research institute, CHAI exists specifically to conduct research rather than combining teaching and investigation like a university department. It is classified as a research organization dedicated to scientific inquiry rather than a membership-based learned society.

## Why It Matters
CHAI represents a critical infrastructure component in the landscape of artificial intelligence development. By organizing as a dedicated research institute rather than a corporate lab or university department, the center provides a focused environment for addressing AI alignment—ensuring that AI systems act in accordance with human intentions.

Research institutes like CHAI are distinct engines of discovery that allow scientists to pursue long-term, high-risk research that might not fit within commercial timelines or instructional obligations. In the context of AI safety, this organizational structure enables the concentrated study of potential risks and safety mechanisms without the distractions of undergraduate education or the profit motives inherent in corporate research.

## Notable For
- **Dedicated AI Safety Focus:** Specifically designated as an "AI safety research center" with a primary focus on "AI alignment."
- **Academic Affiliation:** Officially affiliated with the University of California, Berkeley.
- **Research Institute Classification:** Functions purely as a research organization, distinct from think tanks (which focus on policy) and teaching universities.
- **Digital Presence:** Maintains a dedicated web portal and active social media channels (Twitter/X) established in late 2018.

## Body

### Organizational Classification and Structure
The Center for Human-Compatible Artificial Intelligence is formally classified as both a **research center** and a **research institute**. In the landscape of knowledge-producing institutions, it operates as an entity whose primary purpose is to conduct research. Unlike universities, which balance research with teaching and degree-granting activities, CHAI exists purely to generate new knowledge through systematic investigation.

The organization is differentiated from other similar entities:
- **Distinct from Think Tanks:** While think tanks primarily conduct policy-oriented research to influence government decisions, CHAI focuses on fundamental scientific inquiry in AI alignment.
- **Distinct from Scientific Societies:** Unlike membership-based organizations such as learned societies, CHAI is an operational entity conducting hands-on investigation.
- **Distinct from Corporate Labs:** As a research institute affiliated with a university, it operates within an academic framework rather than a commercial one.

### Location and Affiliation
The center is geographically located in **Berkeley, United States**. It holds a formal affiliation with the **University of California, Berkeley**. This integration with a major public research university provides the center with academic legitimacy and resources while maintaining its specific focus on AI safety.

### History and Timeline
The organization was founded (inception) in **2016**. Its digital footprint on social media platforms like Twitter began on **November 1, 2018**, corresponding to the user ID 1058055843466244096.

### Research Focus: AI Alignment
The defining characteristic of CHAI is its field of work: **AI alignment**. This specialization places the center at the forefront of ensuring that artificial intelligence systems act in ways that are beneficial to humans. This focus on "AI safety" distinguishes it from broader AI research institutes that may focus primarily on capability or efficiency.

### Digital Presence and Identifiers
The center maintains a distinct identity across the web and knowledge systems:
- **Website:** The official portal is located at `http://www.humancompatible.ai/` and is presented in English.
- **Social Media:** The organization operates the handle **@chai_berkeley** on Twitter. As of December 26, 2020, and January 4, 2021, the account had 1,229 followers.
- **Knowledge Graphs:** The entity is tracked by Google with the Knowledge Graph ID `/g/11j1c7dyr4`.
- **Wikidata and Wikipedia:** The entity has a sitelink count of 2, indicating a presence on the English and French Wikipedia editions. It is referenced in Wikidata with provenance attributes confirming its existence and details as of August 2020 and January 2023.

## References

1. Google Knowledge Graph