Mehtaverse

by Swapneel

swapneel_headshot_external.png

Swapneel Mehta

Postdoc at BU and MIT

Founder at SimPPL

Cambridge, MA

Bio

I am a postdoctoral associate at Boston University and MIT conducting research to improve the governance of digital platforms in the age of AI. I founded and co-lead a nonprofit, SimPPL, rebuilding trust on the social internet in 6 countries through responsible AI tools deployed with over 20 newsrooms, nonprofits, civil society orgs., and the UN. We raised ~$350,000 in grants and revenues, and even spun out an award-winning, MIT and UNICEF-incubated tech venture delivering maternal health literacy information in India called Sakhi that aims to reach 5,000 families by 2026.

I received my Ph.D. from NYU Data Science, estimating the causal effects of interventions to improve online safety. I have been an Atlantic Dialogues Emerging Leader, Google Research Innovator, CTS Belfer Fellow, and won several grants from Google, Mozilla, and Wikimedia to support my work. In industry, I worked at Twitter, Adobe, Slack, with Meta at Oxford, (now acquired) startups, and CERN on AI products, algorithms, and applications.


Fostering Digital Trust
I conduct research and develop systems that foster digital trust and credibility, identifying the social, technological, and economic drivers of trust in online information. How do we begin to reverse the consequences of the erosion of digital trust? I study user behavior to understand the longitudinal impact of online harms, design, and evaluate interventions to protect users.
Governing Digital Platforms
At the Platform Governance Lab that I helped set up circa Sept. 2023, we design interventions to limit fake ads without relying on censorship or central authority. I run experiments with over 4,000 online participants in human-AI marketplaces investigating how to limit misleading ads, incentivize the production of honest information, and limit deceptive sales by agentic AI (read more).
Building a Responsible AI Non-profit
At SimPPL, our mission is to rebuild trust on the social internet. We develop societal AI solutions and publish research into AI, platforms, and digital harm. We are a community of 200+ engineers and researchers with 20 active members working with the world's largest newsrooms, nonprofits, the UN, and other agencies in seven countries.
Technology Policy
I am a technologist with experience doing policy research.
Teaching
AI Schools: Since 2017, I have taught students roughly each weekend, initially as part of founding DJ Unicode as an open-source programming bootcamp, expanding it via Unicode Research for collaborative research and recently the NYU AI School for HMGs and non-CS majors to learn applied AI/ML, funded by Deepmind and Genentech, now an in-house program at NYU. I ran a 13-week ML course, UMLSC for undergrads from underserved Indian institutions in 2020 that culminated in the founding of SimPPL, awarded by Google exploreCSR and Mozilla RCC India, winning competitive grants to advance our fellowships program, training undergrads as prolific research engineers.
AI Research: My team has published 15+ papers at ICML, IC2S2, ICWSM, NeurIPS, AAAI, Stanford Trust and Safety, Dg.O, and other workshops/conferences. I've also been a mentor with Lumiere Education, Make a Difference, UAlberta Career Mentoring Program, and actively participate in NYU Alumni Talks to give back to my community.
AI Product Development:I ran our AI Fellowships program and an AI Product Development program across numerous educational institutions through support from Mozilla and Google. Read the outcomes and write to me if you are interested in such a program for your students


Advising: I served as a Community Advisory Board member at the Integrity Institute (read the 2023 Annual Impact), Responsible Tech Affiliate at All Tech is Human, and sit on 3 Boards of Studies at D.J. Sanghvi College of Engineering. I’ve mentored about 130 undergraduates and 4 high-school students. 3 of my mentees are now pursuing their Ph.Ds in Computer Science while several are graduate students, and others work at FAANG companies and startups.


News

selected publications