Insurance Outlook for LLM-Induced Risk

Details

Ressource 1Request a copy Under indefinite embargo.
UNIL restricted access
State: Public
Version: author
License: CC BY 4.0
Serval ID
serval:BIB_12DCF9D6FEF6
Type
A part of a book
Publication sub-type
Chapter: chapter ou part
Collection
Publications
Institution
Title
Insurance Outlook for LLM-Induced Risk
Title of the book
Large Language Models in Cybersecurity
Author(s)
Loïc Maréchal, Daniel Celeny
Publisher
Springer Nature
ISBN
9783031548260
9783031548277
Publication state
Published
Issued date
2024
Peer-reviewed
Oui
Language
english
Abstract
<jats:title>Abstract</jats:title><jats:p>During the development of information systems, security, and safety considerations often take a back seat to market pressures, demanding shorter development cycles, faster releases, and new product features. Unfortunately, right until a cyber-incident, the price of the trade-off between security and safety and other market imperatives is unclear and, given the general rarity of cyber-incidents, often under-estimated. Fortunately, calculating the security and safety side of the trade-off is the domain of expertise of actuaries in insurance companies offering cyber insurances. It used to be an after-thought for most companies since the 2013 Target data breach, which cost nearly 300 million but was covered at 30% by insurance payout. Since then, insurance for risks of information systems malfunctions has become standard for most companies, and premium reduction has become a primary driver for improving cybersecurity costs for companies. The role of this chapter is to transpose what we have learned about the insurance of cyber-incidents over the last couple of decades and use it as a basis to produce a qualitative forecast of the insurance outlook for a security and safety landscape involving LLMs.</jats:p>
Create date
17/07/2024 12:55
Last modification date
18/07/2024 6:06
Usage data