Persona-Conditioned Adversarial Prompting (PCAP): Multi-Identity Red-Teaming for Enhanced Adversarial Prompt Discovery

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Morasso, Cristian, Halimi, Anisa, Hameed, Muhammad Zaid, Leith, Douglas
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Cryptography and Security
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Persona-Conditioned Adversarial Prompting: Multi-Identity Red-Teaming for Adversarial Discovery and Mitigation
di: Morasso, Cristian, et al.
Pubblicazione: (2026)

Towards a Re-evaluation of Data Forging Attacks in Practice
di: Suliman, Mohamed, et al.
Pubblicazione: (2024)

Adversarial Prompt Evaluation: Systematic Benchmarking of Guardrails Against Prompt Input Attacks on LLMs
di: Zizzo, Giulio, et al.
Pubblicazione: (2025)

Autonomous Adversary: Red-Teaming in the age of LLM
di: Mamun, Mohammad, et al.
Pubblicazione: (2026)

Prompt Optimization and Evaluation for LLM Automated Red Teaming
di: Freenor, Michael, et al.
Pubblicazione: (2025)

LoMime: Query-Efficient Membership Inference using Model Extraction in Label-Only Settings
di: Oksuz, Abdullah Caglar, et al.
Pubblicazione: (2026)

AUTOLYCUS: Exploiting Explainable AI (XAI) for Model Extraction Attacks against Interpretable Models
di: Oksuz, Abdullah Caglar, et al.
Pubblicazione: (2023)

Red Team Redemption: A Structured Comparison of Open-Source Tools for Adversary Emulation
di: Landauer, Max, et al.
Pubblicazione: (2024)

Poison Attacks and Adversarial Prompts Against an Informed University Virtual Assistant
di: Fernandez, Ivan A., et al.
Pubblicazione: (2024)

MoJE: Mixture of Jailbreak Experts, Naive Tabular Classifiers as Guard for Prompt Attacks
di: Cornacchia, Giandomenico, et al.
Pubblicazione: (2024)

FlashRT: Towards Computationally and Memory Efficient Red-Teaming for Prompt Injection and Knowledge Corruption
di: Wang, Yanting, et al.
Pubblicazione: (2026)

Enhancing Jailbreak Attacks on LLMs via Persona Prompts
di: Zhang, Zheng, et al.
Pubblicazione: (2025)

PISmith: Reinforcement Learning-based Red Teaming for Prompt Injection Defenses
di: Yin, Chenlong, et al.
Pubblicazione: (2026)

Defending Jailbreak Prompts via In-Context Adversarial Game
di: Zhou, Yujun, et al.
Pubblicazione: (2024)

AutoAdv: Automated Adversarial Prompting for Multi-Turn Jailbreaking of Large Language Models
di: Reddy, Aashray, et al.
Pubblicazione: (2025)

Enhancing Adversarial Transferability with Adversarial Weight Tuning
di: Chen, Jiahao, et al.
Pubblicazione: (2024)

Deciphering the Chaos: Enhancing Jailbreak Attacks via Adversarial Prompt Translation
di: Li, Qizhang, et al.
Pubblicazione: (2024)

Learning-Based Automated Adversarial Red-Teaming for Robustness Evaluation of Large Language Models
di: Wei, Zhang, et al.
Pubblicazione: (2025)

PRM-Free Security Alignment of Large Models via Red Teaming and Adversarial Training
di: Du, Pengfei
Pubblicazione: (2025)

Reflect-Guard: Enhancing LLM Safeguards against Adversarial Prompts via Logical Self-Reflection
di: Lin, Lixing, et al.
Pubblicazione: (2026)

PromptRobust: Towards Evaluating the Robustness of Large Language Models on Adversarial Prompts
di: Zhu, Kaijie, et al.
Pubblicazione: (2023)

Whispers of Wealth: Red-Teaming Google's Agent Payments Protocol via Prompt Injection
di: Debi, Tanusree, et al.
Pubblicazione: (2026)

WARD: Adversarially Robust Defense of Web Agents Against Prompt Injections
di: Cao, Tri, et al.
Pubblicazione: (2026)

Blockchain-Enabled Device-Enhanced Multi-Access Edge Computing in Open Adversarial Environments
di: Islam, Muhammad, et al.
Pubblicazione: (2024)

Measuring the Security of Mobile LLM Agents under Adversarial Prompts from Untrusted Third-Party Channels
di: Du, Chenghao, et al.
Pubblicazione: (2025)

Query-Based Adversarial Prompt Generation
di: Hayase, Jonathan, et al.
Pubblicazione: (2024)

MUZZLE: Adaptive Agentic Red-Teaming of Web Agents Against Indirect Prompt Injection Attacks
di: Syros, Georgios, et al.
Pubblicazione: (2026)

Red Teaming the Mind of the Machine: A Systematic Evaluation of Prompt Injection and Jailbreak Vulnerabilities in LLMs
di: Pathade, Chetan
Pubblicazione: (2025)

Adversarial Attacks on LLM-as-a-Judge Systems: Insights from Prompt Injections
di: Maloyan, Narek, et al.
Pubblicazione: (2025)

Jailbreaking Prompt Attack: A Controllable Adversarial Attack against Diffusion Models
di: Ma, Jiachen, et al.
Pubblicazione: (2024)

A Real-Time, Self-Tuning Moderator Framework for Adversarial Prompt Detection
di: Zhang, Ivan
Pubblicazione: (2025)

LocalAlign: Enabling Generalizable Prompt Injection Defense via Generation of Near-Target Adversarial Examples for Alignment Training
di: Gong, Yuyang, et al.
Pubblicazione: (2026)

How Secure is Secure Code Generation? Adversarial Prompts Put LLM Defenses to the Test
di: Tessa, Melissa, et al.
Pubblicazione: (2026)

SafeReview: Defending LLM-based Review Systems Against Adversarial Hidden Prompts
di: Xin, Yuan, et al.
Pubblicazione: (2026)

Certifying LLM Safety against Adversarial Prompting
di: Kumar, Aounon, et al.
Pubblicazione: (2023)

Sentra-Guard: A Real-Time Multilingual Defense Against Adversarial LLM Prompts
di: Hasan, Md. Mehedi, et al.
Pubblicazione: (2025)

Efficient and Stealthy Jailbreak Attacks via Adversarial Prompt Distillation from LLMs to SLMs
di: Li, Xiang, et al.
Pubblicazione: (2025)

Chain-of-Code Collapse: Reasoning Failures in LLMs via Adversarial Prompting in Code Generation
di: Roh, Jaechul, et al.
Pubblicazione: (2025)

IPI-proxy: An Intercepting Proxy for Red-Teaming Web-Browsing AI Agents Against Indirect Prompt Injection
di: Chia-Pei, et al.
Pubblicazione: (2026)

Improving Privacy Benefits of Redaction
di: Gusain, Vaibhav, et al.
Pubblicazione: (2025)

if (filemtime($filename) + $maxlifetime < time()) { unlink($filename); $count++; } } return $count; } /** * A function that is called internally when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ protected function saveSession($sessId, $data): bool { $sessFile = $this->path . '/sess_' . $sessId; if ($handle = fopen($sessFile, 'w')) { $return = false; // Lock the file for exclusive access to avoid issues with multiple // processes writing session simultaneously: if (flock($handle, LOCK_EX)) { $return = fwrite($handle, $data); // Make sure that there's no trailing data by truncating the file to // the correct length: ftruncate($handle, strlen($data)); fflush($handle); flock($handle, LOCK_UN); } fclose($handle); if ($return !== false) { return true; } } // If we got this far, something went wrong with the file output... // It is tempting to throw an exception here, but this code is called // outside of the context of exception handling, so all we can do is // echo a message.

// Anecdotal testing Today and Yesterday seems to indicate destroy() // is called by the garbage collector and everything is good. // Something to keep in mind though. return 0; } /** * Write function that is called when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ public function write($sessId, $data): bool { if ($this->writesDisabled) { return true; } return $this->saveSession($sessId, $data); } /** * A function that is called internally when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ abstract protected function saveSession($sessId, $data): bool; }

* * @return void */ public function writeClose() { // The assumption is that we're using PHP's ext/session. // session_write_close() will actually overwrite $_SESSION with an // empty array on completion -- which leads to a mismatch between what // is in the storage object and $_SESSION. To get around this, we // temporarily reset $_SESSION to an array, and then re-link it to // the storage object. // // Additionally, while you _can_ write to $_SESSION following a // session_write_close() operation, no changes made to it will be // flushed to the session handler. As such, we now mark the storage // object isImmutable. $storage = $this->getStorage(); if (! $storage->isImmutable()) { $_SESSION = $storage->toArray(true); session_write_close(); $storage->fromArray($_SESSION); $storage->markImmutable(); } } /** * Attempt to set the session name * * If the session has already been started, or if the name provided fails * validation, an exception will be raised. * * @param string $name * @return SessionManager * @throws Exception\InvalidArgumentException */ public function setName($name) { if ($this->sessionExists()) { throw new Exception\InvalidArgumentException( 'Cannot set session name after a session has already started'

Environment & details:

GET Data empty

POST Data empty

Files empty

Key	Value
ui	"standard"
VUFIND_SESSION	"ed76789417e352e3e983bb8391524723"
language	"it"

Session

Key	Value
__Laminas	array:2 [`"_REQUEST_ACCESS_TIME" => 1790867771.4745 "_VALID" => array:1 [ "Laminas\Session\Validator\Id" => "ed76789417e352e3e983bb8391524723" ]`]
SessionState	Laminas\Stdlib\ArrayObject {#938}
FlashMessenger	Laminas\Stdlib\ArrayObject {#1433}
SessionHelper	Laminas\Stdlib\ArrayObject {#1759}

Server/Request Data

Key	Value
REDIRECT_VUFIND_LOCAL_DIR	"/usr/local/vufind/local"
REDIRECT_HTTPS	"on"
REDIRECT_SSL_TLS_SNI	"info.umar.mx"
REDIRECT_STATUS	"200"
VUFIND_LOCAL_DIR	"/usr/local/vufind/local"
HTTPS	"on"
SSL_TLS_SNI	"info.umar.mx"
HTTP_ACCEPT	"/"
HTTP_USER_AGENT	"Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; +claudebot@anthropic.com)"
HTTP_COOKIE	"ui=standard; VUFIND_SESSION=ed76789417e352e3e983bb8391524723; language=it"
HTTP_ACCEPT_ENCODING	"gzip, br, zstd, deflate"
HTTP_HOST	"info.umar.mx"
PATH	"/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/snap/bin"
SERVER_SIGNATURE	"<address>Apache/2.4.66 (Ubuntu) Server at info.umar.mx Port 443</address>\n"
SERVER_SOFTWARE	"Apache/2.4.66 (Ubuntu)"
SERVER_NAME	"info.umar.mx"
SERVER_ADDR	"200.23.222.19"
SERVER_PORT	"443"
REMOTE_ADDR	"216.73.216.147"
DOCUMENT_ROOT	"/usr/local/vufind/public"
REQUEST_SCHEME	"https"
CONTEXT_PREFIX	""
CONTEXT_DOCUMENT_ROOT	"/usr/local/vufind/public"
SERVER_ADMIN	"admin@localhost"
SCRIPT_FILENAME	"/usr/local/vufind/public/index.php"
REMOTE_PORT	"18784"
REDIRECT_URL	""" /Record/\n arxiv_https___arxiv_org_abs_2605_12565/Similar """
GATEWAY_INTERFACE	"CGI/1.1"
SERVER_PROTOCOL	"HTTP/1.1"
REQUEST_METHOD	"GET"
QUERY_STRING	""
REQUEST_URI	"/Record/%0Aarxiv_https___arxiv_org_abs_2605_12565/Similar"
SCRIPT_NAME	"/index.php"
PHP_SELF	"/index.php"
REQUEST_TIME_FLOAT	1790867771.4589
REQUEST_TIME	1790867771

Environment Variables empty

Registered Handlers

0. Whoops\Handler\PrettyPageHandler