Khaoula Chehbouni

Khaoula Chehbouni

PhD Student at McGill University and Mila. I work on fairness and privacy in large language models.

I am a second-year PhD Student in Computer Science at McGill University and Mila under Professor Golnoosh Farnadi and Professor Jackie Chi Kit Cheung. I previously did a Masters in Business Intelligence at HEC Montréal under the co-supervision of Professor Golnoosh Farnadi and Professor Gilles Caporossi. I work on fairness and privacy in machine learning and natural language processing. I am also interested in the broader field of trustworthy machine learning and working with low-resource languages.

News!

  • [January 2025] Our paper “Beyond the Safety Bundle: Auditing the Helpful and Harmless Dataset.” was accepted to the Main Conference at NAACL as an Oral.
  • [January 2025] I was invited as a Panelist by the World Salon for an event in cooperation with KIIT University and Columbia University.
  • [December 2024] Our paper “Enhancing Privacy in the Early Detection of Sexual Predators Through Federated Learning and Differential Privacy.” was accepted to the AAAI Conference - Social Impact Track as an Oral.
  • [December 2024] I completed an AI Policy Certificate with distinctions from the Center for AI and Digital Policy.
  • [November 2024] I was invited as a Panelist at the McGill Technology and Ethics Roundtables.
  • [October 2024] I was invited in the World We Are Building Podcast to talk about my research: Khaoula Chehbouni: Rethinking AI Safety, Hidden Hierarchies in AI Development and Who Gets to Decide What Makes AI Research Worthy of Publication.
  • [October 2024]_ I was invited in the Activists of Tech — The Responsible Tech Podcast to talk about my research and experience in academia: How Existing Safety Mitigation Safeguards Fail in LLMs.
  • [October 2024] Our paper “Beyond the Safety Bundle: Auditing the Helpful and Harmless Dataset.” was accepted in the NeurIPS Workshop Algorithmic Fairness through the Lens of Metrics and Evaluation for a Poster and in the Women in Machine Learning Workshop for an Oral.
  • [October 2024] I was awarded $700USD by Women in Machine Learning and I received $650CA through McGill University’s GREAT Award to attend the NeurIPS Conference.
  • [September 2024] I received the Lorne Trottier Fellowship and was awarded $5000 by McGill University.
  • [August 2024] I successfully passed my PhD Comprehensive Exam.
  • [August 2024] I gave a talk at the AI Policy Compass Event organized by Mila in Montreal called: “Biais et Équité en Intelligence Artificielle”.
  • [August 2024] I attended ACL 2024 in Bangkok and presented our work “From Representational Harms to Quality-of-Service Harms: A Case Study on Llama 2 Safety Safeguards” during a poster session.
  • [June 2024] I gave a talk at the Congrès de la Société de philosophie du Québec called: “Mitigation des risques liés aux Grands Modèles de Langage”.
  • [May 2024] Our paper “From Representational Harms to Quality-of-Service Harms: A Case Study on Llama 2 Safety Safeguards” was accepted for a poster presentation in the Findings of ACL 2024.
  • [May 2024] I started a new position as a senior research scientist at Statistics Canada in their research department to work on fairness and privacy issues in large language models.
  • [May 2024] I won the 3 minute thesis competition at the Responsible AI Track at the Canadian AI Conference at Guelph.
  • [May 2024] I was awarded the prestigious FRQNT Doctoral Training Scholarship. It’s an annual scholarship of \$25000 per year for 4 years (for a total of \$100000).
  • [April 2024] I was awarded one of Mila’s Excellence scholarships for the category EDI in Research, PhD level. It’s an annual scholarship of \$8000 per year for 3 years (for a total of \$24000).
  • [April 2024] Dr. Afaf Taïk and I gave a talk at the Google’s Women Techmaker event called: “Towards Inclusive AI: Understanding bias in machine learning models”.
  • [Feb 2024] I was an expert in a roundtable organized by Mila AI for Humanity about the capabilities, limitations, and risks of GPAIS - from a technical perspective.
  • [Dec 2023] I went to NeurIPS 2023 to present our work in the North Africans in Machine Learning Workshop and I was also a student volunteer during the conference!
  • [Oct 2023] Our survey paper “Leveraging AI for Natural Disaster Management: Takeaways From The Moroccan Earthquake” was accepted for a poster presentation in the Artificial Intelligence for Humanitarian Assistance and Disaster Response Workshop.
  • [Oct 2023] Our project “Unmasking Predators: Safeguarding Vulnerable Moroccan Communities Post-Earthquake” was accepted for a poster presentation in the North Africans in Machine Learning Workshop.
  • [Sep 2023] Our project “Unmasking Predators: Safeguarding Vulnerable Moroccan Communities Post-Earthquake” won the Morocco Solidarity Hackathon.
  • [Sep 2023] I’ve started my PhD at McGill University and Mila under the supervision of Professor Golnoosh Farnadi.
  • [June 2023] I was interviewed by HEC Mag about my Master’s thesis on the detection of sexual predators.
  • [June 2023] I gave a six-hour class about Bias and Fairness in Machine Learning to Statistics Canada employees.
  • [Jan 2023] I’ve started a research internship at Mila under the supervision of Professor Golnoosh Farnadi.
  • [Dec 2022] I attended NeurIPS in New Orleans to present our work “Early Detection of Sexual Predators with Federated Learning” at the International Workshop on Federated Learning and the Women in Machine Learning Workshop.
  • [Oct 2022] Our paper “Early Detection of Sexual Predators with Federated Learning” was accepted at the International Workshop on Federated Learning at NeurIPS.
  • [Oct 2022] Our white paper “AI & Cities: Risks, Applications and Governance” was presented at the Urban Economy Forum.
  • [Sep 2022] My thesis “Mitigating Online Grooming with Federated Learning” won the Best Thesis Award at HEC Montréal for the year 2022.
  • [Sep 2022] I presented our work “Mitigating Online Grooming with Federated Learning” at the Montreal AI Symposium.
  • [Sep 2022] I’ve started working as a Data Scientist in the Methods, Quality, and Research section in Statistics Canada.
  • [March 2022] I’ve completed Ivado’s Action Inclusion Challenge and worked with IBM towards improving women’s representation in STEM.
  • [March 2022] I earned second place in the HEC Data Challenge 2022.
  • [May 2021] I started a data journalism internship at La Presse after winning Ivado’s Data Storytelling Scholarship.