0 / 0
Dangerous use risk for AI
Last updated: Sep 26, 2024
Dangerous use risk for AI
Misuse Icon representing misuse risks.
Risks associated with output
Misuse
New to generative AI

Description

Generative AI models might be used with the sole intention of harming people.

Why is dangerous use a concern for foundation models?

Large language models are often trained on vast amounts of publicly-available information that may include information on harming others. A model that has this potential must be carefully evaluated for such content and properly governed.

Background image for risks associated with input
Example

AI-based Cyberattacks

According to the source article, hackers are increasingly experimenting with ChatGPT and other AI tools, enabling a wider range of actors to carry out cyberattacks and scams. Microsoft has warned that state-backed hackers have been using OpenAI’s LLMs to improve their cyberattacks, refining scripts, and improve their targeted techniques. The article also mentions about a case where Microsoft and OpenAI say they detected attempts from attackers and sharp increase in cyberattacks targeting government offices.

Background image for risks associated with input
Example

AI-based Biological Attacks

As per the source article, large language models could help in the planning and execution of a biological attack. Several test scenarios are mentioned such as using LLMs to identify biological agents and their relative chances of harm to human life. The article also highlighted the open question which is the level of threat LLMs present beyond the harmful information that is readily available online.

Parent topic: AI risk atlas

We provide examples covered by the press to help explain many of the foundation models' risks. Many of these events covered by the press are either still evolving or have been resolved, and referencing them can help the reader understand the potential risks and work towards mitigations. Highlighting these examples are for illustrative purposes only.

Generative AI search and answer
These answers are generated by a large language model in watsonx.ai based on content from the product documentation. Learn more