A cautionary (Canterbury) tale of causal AI

Christopher Hand

A cautionary (Canterbury) tale of causal AI

Research output: Contribution to conference › Paper › peer-review

Abstract

The development paper investigates claims made for Causal AI methods - specifically that they can identify underlying causal relationships and not confuse correlation with causation. We put this to the test by creating a dataset containing both causal and correaltional relationships and applying a number of Causal AI algorithms. All of them identified the correlational relationship as causal. We highlight that caution, and subject knowledge, are needed when interpreting the output produced.

Original language	English
Publication status	Published - 4 Sept 2025
Event	British Academy of Management 2025 Conference (BAM2025) - University of Kent, Canterbury, United Kingdom Duration: 1 Sept 2025 → 5 Sept 2025 Conference number: 39 https://www.bam.ac.uk/events-landing/past-conferences/2025-conference.html

Conference

Conference	British Academy of Management 2025 Conference (BAM2025)
Country/Territory	United Kingdom
City	Canterbury
Period	1/09/25 → 5/09/25
Internet address	https://www.bam.ac.uk/events-landing/past-conferences/2025-conference.html

Cite this

@conference{52f16f6ae54247cabed1c10c627f46ab,

title = "A cautionary (Canterbury) tale of causal AI",

abstract = "The development paper investigates claims made for Causal AI methods - specifically that they can identify underlying causal relationships and not confuse correlation with causation. We put this to the test by creating a dataset containing both causal and correaltional relationships and applying a number of Causal AI algorithms. All of them identified the correlational relationship as causal. We highlight that caution, and subject knowledge, are needed when interpreting the output produced. ",

author = "Christopher Hand",

year = "2025",

month = sep,

day = "4",

language = "English",

note = "British Academy of Management 2025 Conference (BAM2025) ; Conference date: 01-09-2025 Through 05-09-2025",

url = "https://www.bam.ac.uk/events-landing/past-conferences/2025-conference.html ",

}

TY - CONF

T1 - A cautionary (Canterbury) tale of causal AI

AU - Hand, Christopher

N1 - Conference code: 39

PY - 2025/9/4

Y1 - 2025/9/4

N2 - The development paper investigates claims made for Causal AI methods - specifically that they can identify underlying causal relationships and not confuse correlation with causation. We put this to the test by creating a dataset containing both causal and correaltional relationships and applying a number of Causal AI algorithms. All of them identified the correlational relationship as causal. We highlight that caution, and subject knowledge, are needed when interpreting the output produced.

AB - The development paper investigates claims made for Causal AI methods - specifically that they can identify underlying causal relationships and not confuse correlation with causation. We put this to the test by creating a dataset containing both causal and correaltional relationships and applying a number of Causal AI algorithms. All of them identified the correlational relationship as causal. We highlight that caution, and subject knowledge, are needed when interpreting the output produced.

M3 - Paper

T2 - British Academy of Management 2025 Conference (BAM2025)

Y2 - 1 September 2025 through 5 September 2025

ER -

A cautionary (Canterbury) tale of causal AI

Abstract

Conference

Fingerprint

Cite this