DecipherGuard: Understanding and Deciphering Jailbreak Prompts for a Safer Deployment of Intelligent Software Systems Article Swipe
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2509.16870
Intelligent software systems powered by Large Language Models (LLMs) are increasingly deployed in critical sectors, raising concerns about their safety during runtime. Through an industry-academic collaboration when deploying an LLM-powered virtual customer assistant, a critical software engineering challenge emerged: how to enhance a safer deployment of LLM-powered software systems at runtime? While LlamaGuard, the current state-of-the-art runtime guardrail, offers protection against unsafe inputs, our study reveals a Defense Success Rate (DSR) drop of 24% under obfuscation- and template-based jailbreak attacks. In this paper, we propose DecipherGuard, a novel framework that integrates a deciphering layer to counter obfuscation-based prompts and a low-rank adaptation mechanism to enhance guardrail effectiveness against template-based attacks. Empirical evaluation on over 22,000 prompts demonstrates that DecipherGuard improves DSR by 36% to 65% and Overall Guardrail Performance (OGP) by 20% to 50% compared to LlamaGuard and two other runtime guardrails. These results highlight the effectiveness of DecipherGuard in defending LLM-powered software systems against jailbreak attacks during runtime.
Related Topics
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2509.16870
- https://arxiv.org/pdf/2509.16870
- OA Status
- green
- OpenAlex ID
- https://openalex.org/W4415252640
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4415252640Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2509.16870Digital Object Identifier
- Title
-
DecipherGuard: Understanding and Deciphering Jailbreak Prompts for a Safer Deployment of Intelligent Software SystemsWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2025Year of publication
- Publication date
-
2025-09-21Full publication date if available
- Authors
-
Rui Yang, Michael C. Fu, Chakkrit Tantithamthavorn, Chetan Arora, Gunel Gulmammadova, Joey ChuaList of authors in order
- Landing page
-
https://arxiv.org/abs/2509.16870Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2509.16870Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2509.16870Direct OA link when available
- Cited by
-
0Total citation count in OpenAlex
Full payload
| id | https://openalex.org/W4415252640 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2509.16870 |
| ids.doi | https://doi.org/10.48550/arxiv.2509.16870 |
| ids.openalex | https://openalex.org/W4415252640 |
| fwci | |
| type | preprint |
| title | DecipherGuard: Understanding and Deciphering Jailbreak Prompts for a Safer Deployment of Intelligent Software Systems |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T10734 |
| topics[0].field.id | https://openalex.org/fields/17 |
| topics[0].field.display_name | Computer Science |
| topics[0].score | 0.986299991607666 |
| topics[0].domain.id | https://openalex.org/domains/3 |
| topics[0].domain.display_name | Physical Sciences |
| topics[0].subfield.id | https://openalex.org/subfields/1710 |
| topics[0].subfield.display_name | Information Systems |
| topics[0].display_name | Information and Cyber Security |
| topics[1].id | https://openalex.org/T11241 |
| topics[1].field.id | https://openalex.org/fields/17 |
| topics[1].field.display_name | Computer Science |
| topics[1].score | 0.9484999775886536 |
| topics[1].domain.id | https://openalex.org/domains/3 |
| topics[1].domain.display_name | Physical Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/1711 |
| topics[1].subfield.display_name | Signal Processing |
| topics[1].display_name | Advanced Malware Detection Techniques |
| topics[2].id | https://openalex.org/T12034 |
| topics[2].field.id | https://openalex.org/fields/17 |
| topics[2].field.display_name | Computer Science |
| topics[2].score | 0.939300000667572 |
| topics[2].domain.id | https://openalex.org/domains/3 |
| topics[2].domain.display_name | Physical Sciences |
| topics[2].subfield.id | https://openalex.org/subfields/1710 |
| topics[2].subfield.display_name | Information Systems |
| topics[2].display_name | Digital and Cyber Forensics |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2509.16870 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4306400194 |
| locations[0].source.issn | |
| locations[0].source.type | repository |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | |
| locations[0].source.is_core | False |
| locations[0].source.is_in_doaj | False |
| locations[0].source.display_name | arXiv (Cornell University) |
| locations[0].source.host_organization | https://openalex.org/I205783295 |
| locations[0].source.host_organization_name | Cornell University |
| locations[0].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[0].license | |
| locations[0].pdf_url | https://arxiv.org/pdf/2509.16870 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | text |
| locations[0].license_id | |
| locations[0].is_accepted | False |
| locations[0].is_published | False |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | http://arxiv.org/abs/2509.16870 |
| locations[1].id | doi:10.48550/arxiv.2509.16870 |
| locations[1].is_oa | True |
| locations[1].source.id | https://openalex.org/S4306400194 |
| locations[1].source.issn | |
| locations[1].source.type | repository |
| locations[1].source.is_oa | True |
| locations[1].source.issn_l | |
| locations[1].source.is_core | False |
| locations[1].source.is_in_doaj | False |
| locations[1].source.display_name | arXiv (Cornell University) |
| locations[1].source.host_organization | https://openalex.org/I205783295 |
| locations[1].source.host_organization_name | Cornell University |
| locations[1].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[1].license | cc-by |
| locations[1].pdf_url | |
| locations[1].version | |
| locations[1].raw_type | article |
| locations[1].license_id | https://openalex.org/licenses/cc-by |
| locations[1].is_accepted | False |
| locations[1].is_published | |
| locations[1].raw_source_name | |
| locations[1].landing_page_url | https://doi.org/10.48550/arxiv.2509.16870 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5039983188 |
| authorships[0].author.orcid | https://orcid.org/0000-0002-3209-0456 |
| authorships[0].author.display_name | Rui Yang |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Yang, Rui |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5102710465 |
| authorships[1].author.orcid | https://orcid.org/0000-0001-7211-3491 |
| authorships[1].author.display_name | Michael C. Fu |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Fu, Michael |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5081449581 |
| authorships[2].author.orcid | https://orcid.org/0000-0002-5516-9984 |
| authorships[2].author.display_name | Chakkrit Tantithamthavorn |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Tantithamthavorn, Chakkrit |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5019739552 |
| authorships[3].author.orcid | https://orcid.org/0000-0003-1466-7386 |
| authorships[3].author.display_name | Chetan Arora |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Arora, Chetan |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5119812678 |
| authorships[4].author.orcid | |
| authorships[4].author.display_name | Gunel Gulmammadova |
| authorships[4].author_position | middle |
| authorships[4].raw_author_name | Gulmammadova, Gunel |
| authorships[4].is_corresponding | False |
| authorships[5].author.id | https://openalex.org/A5116748729 |
| authorships[5].author.orcid | |
| authorships[5].author.display_name | Joey Chua |
| authorships[5].author_position | last |
| authorships[5].raw_author_name | Chua, Joey |
| authorships[5].is_corresponding | False |
| has_content.pdf | False |
| has_content.grobid_xml | False |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://arxiv.org/pdf/2509.16870 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-10-16T00:00:00 |
| display_name | DecipherGuard: Understanding and Deciphering Jailbreak Prompts for a Safer Deployment of Intelligent Software Systems |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T06:51:31.235846 |
| primary_topic.id | https://openalex.org/T10734 |
| primary_topic.field.id | https://openalex.org/fields/17 |
| primary_topic.field.display_name | Computer Science |
| primary_topic.score | 0.986299991607666 |
| primary_topic.domain.id | https://openalex.org/domains/3 |
| primary_topic.domain.display_name | Physical Sciences |
| primary_topic.subfield.id | https://openalex.org/subfields/1710 |
| primary_topic.subfield.display_name | Information Systems |
| primary_topic.display_name | Information and Cyber Security |
| cited_by_count | 0 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2509.16870 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4306400194 |
| best_oa_location.source.issn | |
| best_oa_location.source.type | repository |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | |
| best_oa_location.source.is_core | False |
| best_oa_location.source.is_in_doaj | False |
| best_oa_location.source.display_name | arXiv (Cornell University) |
| best_oa_location.source.host_organization | https://openalex.org/I205783295 |
| best_oa_location.source.host_organization_name | Cornell University |
| best_oa_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| best_oa_location.license | |
| best_oa_location.pdf_url | https://arxiv.org/pdf/2509.16870 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | text |
| best_oa_location.license_id | |
| best_oa_location.is_accepted | False |
| best_oa_location.is_published | False |
| best_oa_location.raw_source_name | |
| best_oa_location.landing_page_url | http://arxiv.org/abs/2509.16870 |
| primary_location.id | pmh:oai:arXiv.org:2509.16870 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4306400194 |
| primary_location.source.issn | |
| primary_location.source.type | repository |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | |
| primary_location.source.is_core | False |
| primary_location.source.is_in_doaj | False |
| primary_location.source.display_name | arXiv (Cornell University) |
| primary_location.source.host_organization | https://openalex.org/I205783295 |
| primary_location.source.host_organization_name | Cornell University |
| primary_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| primary_location.license | |
| primary_location.pdf_url | https://arxiv.org/pdf/2509.16870 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | text |
| primary_location.license_id | |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | http://arxiv.org/abs/2509.16870 |
| publication_date | 2025-09-21 |
| publication_year | 2025 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 33, 42, 66, 86, 91, 99 |
| abstract_inverted_index.In | 80 |
| abstract_inverted_index.an | 23, 28 |
| abstract_inverted_index.at | 49 |
| abstract_inverted_index.by | 4, 121, 130 |
| abstract_inverted_index.in | 12, 149 |
| abstract_inverted_index.of | 45, 72, 147 |
| abstract_inverted_index.on | 112 |
| abstract_inverted_index.to | 40, 94, 103, 123, 132, 135 |
| abstract_inverted_index.we | 83 |
| abstract_inverted_index.20% | 131 |
| abstract_inverted_index.24% | 73 |
| abstract_inverted_index.36% | 122 |
| abstract_inverted_index.50% | 133 |
| abstract_inverted_index.65% | 124 |
| abstract_inverted_index.DSR | 120 |
| abstract_inverted_index.and | 76, 98, 125, 137 |
| abstract_inverted_index.are | 9 |
| abstract_inverted_index.how | 39 |
| abstract_inverted_index.our | 63 |
| abstract_inverted_index.the | 53, 145 |
| abstract_inverted_index.two | 138 |
| abstract_inverted_index.Rate | 69 |
| abstract_inverted_index.drop | 71 |
| abstract_inverted_index.over | 113 |
| abstract_inverted_index.that | 89, 117 |
| abstract_inverted_index.this | 81 |
| abstract_inverted_index.when | 26 |
| abstract_inverted_index.(DSR) | 70 |
| abstract_inverted_index.(OGP) | 129 |
| abstract_inverted_index.Large | 5 |
| abstract_inverted_index.These | 142 |
| abstract_inverted_index.While | 51 |
| abstract_inverted_index.about | 17 |
| abstract_inverted_index.layer | 93 |
| abstract_inverted_index.novel | 87 |
| abstract_inverted_index.other | 139 |
| abstract_inverted_index.safer | 43 |
| abstract_inverted_index.study | 64 |
| abstract_inverted_index.their | 18 |
| abstract_inverted_index.under | 74 |
| abstract_inverted_index.(LLMs) | 8 |
| abstract_inverted_index.22,000 | 114 |
| abstract_inverted_index.Models | 7 |
| abstract_inverted_index.during | 20, 157 |
| abstract_inverted_index.offers | 58 |
| abstract_inverted_index.paper, | 82 |
| abstract_inverted_index.safety | 19 |
| abstract_inverted_index.unsafe | 61 |
| abstract_inverted_index.Defense | 67 |
| abstract_inverted_index.Overall | 126 |
| abstract_inverted_index.Success | 68 |
| abstract_inverted_index.Through | 22 |
| abstract_inverted_index.against | 60, 107, 154 |
| abstract_inverted_index.attacks | 156 |
| abstract_inverted_index.counter | 95 |
| abstract_inverted_index.current | 54 |
| abstract_inverted_index.enhance | 41, 104 |
| abstract_inverted_index.inputs, | 62 |
| abstract_inverted_index.powered | 3 |
| abstract_inverted_index.prompts | 97, 115 |
| abstract_inverted_index.propose | 84 |
| abstract_inverted_index.raising | 15 |
| abstract_inverted_index.results | 143 |
| abstract_inverted_index.reveals | 65 |
| abstract_inverted_index.runtime | 56, 140 |
| abstract_inverted_index.systems | 2, 48, 153 |
| abstract_inverted_index.virtual | 30 |
| abstract_inverted_index.Language | 6 |
| abstract_inverted_index.attacks. | 79, 109 |
| abstract_inverted_index.compared | 134 |
| abstract_inverted_index.concerns | 16 |
| abstract_inverted_index.critical | 13, 34 |
| abstract_inverted_index.customer | 31 |
| abstract_inverted_index.deployed | 11 |
| abstract_inverted_index.emerged: | 38 |
| abstract_inverted_index.improves | 119 |
| abstract_inverted_index.low-rank | 100 |
| abstract_inverted_index.runtime. | 21, 158 |
| abstract_inverted_index.runtime? | 50 |
| abstract_inverted_index.sectors, | 14 |
| abstract_inverted_index.software | 1, 35, 47, 152 |
| abstract_inverted_index.Empirical | 110 |
| abstract_inverted_index.Guardrail | 127 |
| abstract_inverted_index.challenge | 37 |
| abstract_inverted_index.defending | 150 |
| abstract_inverted_index.deploying | 27 |
| abstract_inverted_index.framework | 88 |
| abstract_inverted_index.guardrail | 105 |
| abstract_inverted_index.highlight | 144 |
| abstract_inverted_index.jailbreak | 78, 155 |
| abstract_inverted_index.mechanism | 102 |
| abstract_inverted_index.LlamaGuard | 136 |
| abstract_inverted_index.adaptation | 101 |
| abstract_inverted_index.assistant, | 32 |
| abstract_inverted_index.deployment | 44 |
| abstract_inverted_index.evaluation | 111 |
| abstract_inverted_index.guardrail, | 57 |
| abstract_inverted_index.integrates | 90 |
| abstract_inverted_index.protection | 59 |
| abstract_inverted_index.Intelligent | 0 |
| abstract_inverted_index.LLM-powered | 29, 46, 151 |
| abstract_inverted_index.LlamaGuard, | 52 |
| abstract_inverted_index.Performance | 128 |
| abstract_inverted_index.deciphering | 92 |
| abstract_inverted_index.engineering | 36 |
| abstract_inverted_index.guardrails. | 141 |
| abstract_inverted_index.demonstrates | 116 |
| abstract_inverted_index.increasingly | 10 |
| abstract_inverted_index.obfuscation- | 75 |
| abstract_inverted_index.DecipherGuard | 118, 148 |
| abstract_inverted_index.collaboration | 25 |
| abstract_inverted_index.effectiveness | 106, 146 |
| abstract_inverted_index.DecipherGuard, | 85 |
| abstract_inverted_index.template-based | 77, 108 |
| abstract_inverted_index.state-of-the-art | 55 |
| abstract_inverted_index.industry-academic | 24 |
| abstract_inverted_index.obfuscation-based | 96 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 6 |
| citation_normalized_percentile |