iBet uBet web content aggregator. Adding the entire web to your favor.
iBet uBet web content aggregator. Adding the entire web to your favor.



Link to original content: https://doi.org/10.48550/arXiv.2304.05197
[2304.05197] Multi-step Jailbreaking Privacy Attacks on ChatGPT
Skip to main content
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arXiv logo
Cornell University Logo

Computer Science > Computation and Language

arXiv:2304.05197 (cs)
[Submitted on 11 Apr 2023 (v1), last revised 1 Nov 2023 (this version, v3)]

Title:Multi-step Jailbreaking Privacy Attacks on ChatGPT

Authors:Haoran Li, Dadi Guo, Wei Fan, Mingshi Xu, Jie Huang, Fanpu Meng, Yangqiu Song
View PDF
Abstract:With the rapid progress of large language models (LLMs), many downstream NLP tasks can be well solved given appropriate prompts. Though model developers and researchers work hard on dialog safety to avoid generating harmful content from LLMs, it is still challenging to steer AI-generated content (AIGC) for the human good. As powerful LLMs are devouring existing text data from various domains (e.g., GPT-3 is trained on 45TB texts), it is natural to doubt whether the private information is included in the training data and what privacy threats can these LLMs and their downstream applications bring. In this paper, we study the privacy threats from OpenAI's ChatGPT and the New Bing enhanced by ChatGPT and show that application-integrated LLMs may cause new privacy threats. To this end, we conduct extensive experiments to support our claims and discuss LLMs' privacy implications.
Comments: Findings of EMNLP 2023. Updated with results on open-source LLMs
Subjects: Computation and Language (cs.CL); Cryptography and Security (cs.CR)
Cite as: arXiv:2304.05197 [cs.CL]
  (or arXiv:2304.05197v3 [cs.CL] for this version)
  https://doi.org/10.48550/arXiv.2304.05197

Submission history

From: Haoran Li [view email]
[v1] Tue, 11 Apr 2023 13:05:04 UTC (7,641 KB)
[v2] Thu, 18 May 2023 17:11:40 UTC (7,987 KB)
[v3] Wed, 1 Nov 2023 05:14:01 UTC (7,989 KB)
Full-text links:

Access Paper:

view license
Current browse context:
cs.CL
< prev   |   next >
new | recent | 2023-04
Change to browse by:
cs
cs.CR

References & Citations

export BibTeX citation

Bookmark

BibSonomy logo Reddit logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)