x

LESSWRONG
LW

AI Misuse — LessWrong

AI Misuse

Edited by Raemon last updated 1st May 2023

AI misuse. Humans using AI in a way that harms humanity.

1

1

Posts tagged AI Misuse

2

63Managing catastrophic misuse without robust AIs

ryan_greenblatt, Buck

2y

17

2

30Adversarial Robustness Could Help Prevent Catastrophic Misuse

2y

18

2

17Jailbreaking AI models to Phish Elderly Victims

Simon Lermen, Fred Heiding

16d

0

2

17Distinguishing misuse is difficult and uncomfortable

3y

3

1

94Covert Malicious Finetuning

Tony Wang, dannyhalawi

1y

4

1

81Human study on AI spear phishing campaigns

Simon Lermen, Fred Heiding, Andrew Kao

11mo

8

1

38Scalable And Transferable Black-Box Jailbreaks For Language Models Via Persona Modulation

Soroush Pour, rusheb, Quentin FEUILLADE--MONTIXI, Arush, scasper

2y

2

1

23On excluding dangerous information from training

2y

5

1

22Proposal: we should start referring to the risk from unaligned AI as a type of *accident risk*

Christopher King

3y

6

1

18Proposal: Align Systems Earlier In Training

3y

0

1

7AI and Biological Risk: Forecasting Key Capability Thresholds

Alvin Ånestrand

2mo

4

1

6How to solve the misuse problem assuming that in 10 years the default scenario is that AGI agents are capable of synthetizing pathogens

1y

0

1

4Visual Prompt Injections: Results on testing AI spam-defense and AI vulnerability to deceptive web ads.

6mo

0

1

3Misalignment or misuse? The AGI alignment tradeoff

5mo

0

1

2Technical Risks of (Lethal) Autonomous Weapons Systems

1y

0

Load More (15/15)

Add Posts