Mohamed!

I'm a cognitive science student at the University of Toronto. I'm also a research assistant at the Citizen Lab and in the Department of Computer Science's HCI group. My research interests lie at the intersection of technology and human rights. Below you can find some of my work.

Papers

An Analysis of Chinese Censorship Bias in LLMs
(To Appear) Proceedings on Privacy Enhancing Technogies (PoPETs 2025), Issue 4
Mohamed Ahmed, Jeffrey Knockel, and Rachel Greenstadt

When a large language model (LLM) has been trained on text featuring social biases, those biases implicitly impact the outputs of the model. Training an LLM on sanitized content, i.e., those pieces of content which remain after being subjected to state censorship (including alterations, deletions, and self-imposed censorship), results in what we term censorship bias. A model impacted by censorship bias may be less likely to reflect views that are routinely prohibited and more likely to reflect views that are not. This may particularly be an issue when interfacing with a model in a language that is predominantly used in a region with strong censorship laws. In this work, we outline what censorship bias is, introduce a novel methodology for identifying and measuring it, and apply that methodology to evaluate the most popular current LLMs. As part of the contributions of this work we designed and evaluated CensorshipDetector, a Chinese language text classification model which we use as part of our experimental design. Our evaluation of CensorshipDetector found it to be 91% accurate at differentiating between sanitized content and non-sanitized content. Our testing revealed evidence of censorship bias across all of the models we evaluated. Finally, we outline the potential harms of censorship bias, namely the exportation of information manipulation that would have primarily harmed a domestic audience to diaspora, as well as recommendations to various stakeholders to limit the harms of censorship bias and prevent it in the future.

Compassionately: Increasing Plurality Awareness Through Community-powered AI
Muslims in ML Workshop, co-located with NeurIPS 2024
Hala Sheta, Mohamed Ahmed, and Syed Ishtiaque Ahmed

The prevalence of Islamophobia has resulted in the continual discrimination of Muslims in North America, where polls have found them to be viewed in a negative light, specifically as a rigid, violent and monolithic community. Furthermore, within the Muslim community, many are ignorant and intolerant of differing perspectives to their own beliefs, viewing Islamic jurisprudence as a one-sided, static truth. However, recent work in this domain rightfully reframe it as a human-centered field, and work to amplify minority voices (e.g. women) that are often undermined in textual interpretations. So, our research is centered around utilizing community-powered AI to increase plurality awareness and understanding, both within and outside of the Muslim community. The end goal is to develop a semi-automated hate-speech detection system that educates its users about the multiplicity of perspectives surrounding a topic of interest. This paper will serve as foundational work to this overarching goal by gauging the current climate surrounding the discussion of Islamic perspectives both through a qualitative analysis of online discussions on Reddit and conducting a controlled user study.

Banned Books: Analysis of Censorship on Amazon.com
Citizen Lab Report
Jeffrey Knockel, Jakub Dalek, Noura Aljizawi, Mohamed Ahmed, Levi Meletti, and Justin Lau

We analyze the system Amazon deploys on the US “amazon.com” storefront to restrict shipments of certain products to specific regions. We found 17,050 products that Amazon restricted from being shipped to at least one world region. While many of the shipping restrictions are related to regulations involving WiFi, car seats, and other heavily regulated product categories, the most common product category restricted by Amazon in our study was books.

Extended Abstract: The Impact of Online Censorship on LLMs
Free and Open Communications on the Internet (FOCI 2024)
Mohamed Ahmed, and Jeffrey Knockel

While there has been a substantial and growing effort to identify, analyze, and mitigate implicit biases in large language models (LLMs), little emphasis has been put on measuring the impacts of online censorship practices on these models. In the same way that biases and false information ingrained in training datasets may manifest themselves implicitly in the outputs of models, training a generative model on censored content, i.e., content that is subject to censorship rules, may impact the views reflected in its responses. Furthermore, the issue may be accentuated for speakers of a language if a substantial portion of a model’s training data in that language has been subject to censorship. We propose an experiment for analyzing the effects of online censorship on black-box LLMs by evaluating models' responses to prompts made in Simplified and Traditional Chinese to determine their similarity to known censored content. Our exploratory testing suggests that, when asked in Simplified Chinese, LLMs provide answers largely in keeping with Chinese information control requirements unlike when asked in Traditional Chinese. Due to the popular usage of LLMs globally, we hypothesize that LLMs unwittingly export information manipulation that would have primarily harmed a domestic audience to diaspora and other Chinese speakers living abroad.

Talks

An Analysis of Chinese Censorship Bias in LLMs
Privacy Enhancing Technologies Symposium (PETS 2025)
Mohamed Ahmed

Reframing AI Alignment in Terms of Censorship and Privacy
HotPETs 2025
Mohamed Ahmed

Compassionately: Increasing Plurality Awareness Through Community-powered AI
Muslims in ML Workshop, co-located with NeurIPS 2024
Hala Sheta, and Mohamed Ahmed

The Impact of Online Censorship on LLMs
Free and Open Communications on the Internet (FOCI 2024)
Mohamed Ahmed

Hi, I'm Mohamed :)

Papers

Talks

Posters