publications

2026

1. Under Review

   Thought-Transfer: Indirect Targeted Poisoning Attacks on Chain-of-Thought Reasoning Models

   Harsh Chaudhari, Ethan Rathbun , Hanna Foerster , and 5 more authors

   arXiv, 2026

   Bib PDF

   @article{poisoningreasoningmodels26,
     title = {Thought-Transfer: Indirect Targeted Poisoning Attacks on Chain-of-Thought Reasoning Models},
     author = {Chaudhari, Harsh and Rathbun, Ethan and Foerster, Hanna and Hayes, Jamie and Jagielski, Matthew and Nasr, Milad and Shumailov, Ilia and Oprea, Alina},
     journal = {arXiv},
     year = {2026},
   }

2. IEEE S&P

   Exploiting Leaderboards for Large-Scale Distribution of Malicious Generative Models

   Anshuman Suri* , Harsh Chaudhari*, Yuefeng Peng* , and 3 more authors

   IEEE Security & Privacy, 2026

   Bib PDF

   @article{poisonleaderboard25,
     title = {Exploiting Leaderboards for Large-Scale Distribution of Malicious Generative Models},
     author = {Suri*, Anshuman and Chaudhari*, Harsh and Peng*, Yuefeng and Naseh*, Ali and Oprea, Alina and Houmansadr, Amir},
     journal = {IEEE Security & Privacy},
     year = {2026},
   }

3. TAISAP

   Phantom: General Backdoor Attacks on Retrieval Augmented Language Generation

   Harsh Chaudhari*, Giorgio Severi* , John Abascal , and 5 more authors

   ACM Transactions on AI Security and Privacy, 2026

   Bib PDF Code

   @article{phantom,
     title = {Phantom: General Backdoor Attacks on Retrieval Augmented Language Generation},
     author = {Chaudhari*, Harsh and Severi*, Giorgio and Abascal, John and Suri, Anshuman and Jagielski, Matthew and Choquette-Choo, Christopher and Nasr, Milad and Oprea, Alina},
     journal = {ACM Transactions on AI Security and Privacy},
     year = {2026},
   }

2025

1. CCS

   Cascading Adversarial Bias from Injection to Distillation in Language Models

   Harsh Chaudhari, Jamie Hayes , Matthew Jagielski , and 3 more authors

   ACM CCS, 2025

   Bib PDF

   @article{cascading_adversarial_bias25,
     title = {Cascading Adversarial Bias from Injection to Distillation in Language Models},
     author = {Chaudhari, Harsh and Hayes, Jamie and Jagielski, Matthew and Shumailov, Ilia and Nasr, Milad and Oprea, Alina},
     journal = {ACM CCS},
     year = {2025},
   }

2024

1. ICLR

   CHAMELEON: Increasing Label-Only Membership Inference Attacks with Adaptive Poisoning

   Harsh Chaudhari, Giorgio Severi , Alina Oprea , and 1 more author

   ICLR, 2024

   Bib PDF Code

   @article{chameleon24,
     title = {CHAMELEON: Increasing Label-Only Membership Inference Attacks with Adaptive Poisoning},
     author = {Chaudhari, Harsh and Severi, Giorgio and Oprea, Alina and Ullman, Jonathan},
     journal = {ICLR},
     year = {2024},
   }

2023

1. IEEE S&P

   SNAP: Efficient Extraction of Private Properties with Poisoning

   Harsh Chaudhari, John Abascal , Alina Oprea , and 3 more authors

   IEEE Security & Privacy, 2023

   Bib PDF Code

   @article{snap23,
     title = {SNAP: Efficient Extraction of Private Properties with Poisoning},
     author = {Chaudhari, Harsh and Abascal, John and Oprea, Alina and Jagielski, Matthew and Tramer, Florian and Ullman, Jonathan},
     journal = {IEEE Security & Privacy},
     year = {2023},
   }

2. SATML

   SafeNet: The Unreasonable Effectiveness of Ensembles in Private Collaborative Learning

   Harsh Chaudhari, Matthew Jagielski , and Alina Oprea

   IEEE SATML, 2023

   Bib PDF

   @article{satml23,
     title = {SafeNet: The Unreasonable Effectiveness of Ensembles in Private Collaborative Learning},
     author = {Chaudhari, Harsh and Jagielski, Matthew and Oprea, Alina},
     journal = {IEEE SATML},
     year = {2023},
   }

2020

1. PETS

   FLASH: Fast and Robust Framework for Privacy-preserving Machine Learning

   Megha Byali , Harsh Chaudhari, Arpita Patra , and 1 more author

   PETS, 2020

   Bib PDF

   @article{flash20,
     title = {FLASH: Fast and Robust Framework for Privacy-preserving Machine Learning},
     author = {Byali, Megha and Chaudhari, Harsh and Patra, Arpita and Suresh, Ajith},
     journal = {PETS},
     year = {2020},
   }

2. NDSS

   Trident: Efficient 4PC Framework for Privacy Preserving Machine Learning

   Harsh Chaudhari, Rahul Rachuri , and Ajith Suresh

   NDSS, 2020

   Bib PDF

   @article{trident20,
     title = {Trident: Efficient 4PC Framework for Privacy Preserving Machine Learning},
     author = {Chaudhari, Harsh and Rachuri, Rahul and Suresh, Ajith},
     journal = {NDSS},
     year = {2020},
   }