In AI, alignment is the goal. Steerability is how you get there
Q & A
Kim Martineau
Fairness, Accountability, Transparency
Biases can lead to systematic disadvantages for marginalized individuals and groups — and they can arise in any point in the AI development lifecycle. To increase the accountability of high-risk AI systems, we're developing technologies to increase their end-to-end transparency and fairness.
Our work
IBM’s safety checkers top a new AI benchmark
NewsKim MartineauIBM’s Mikhail Yurochkin wants to make AI’s “cool” factor tangible
ResearchKim MartineauIBM reaffirms its commitment to the Rome Call for AI ethics
NewsMike MurphyWhat is red teaming for generative AI?
ExplainerKim MartineauThe latest AI safety method is a throwback to our maritime past
ResearchKim Martineau- See more of our work on Fairness, Accountability, Transparency
Projects
We're developing technological solutions to assist subject matter experts with their scientific workflows by enabling the Human-AI co-creation process.
Publications
Highlight All the Phrases: Enhancing LLM Transparency through Visual Factuality Indicators
- Hyo Jin Do
- Rachel Ostrand
- et al.
- 2025
- AIES 2025
Bridging Expertise and Participation in AI: Multistakeholder Approaches to Safer AI Systems for Youth Online Safety
- Ozioma Collins Oguine
- Johanna Olesk
- et al.
- 2025
- CSCW 2025
3rd TrustAI Workshop: Building Public Awareness and Engagement
- Miriam Rateike
- Brian Mboya
- et al.
- 2025
- DLI 2025
DECASTE: Unveiling Caste Stereotypes in Large Language Models through Multi-Dimensional Bias Analysis
- Prashanth Vijayaraghavan
- Soroush Vosoughi
- et al.
- 2025
- IJCAI 2025
Protecting Users From Themselves: Safeguarding Contextual Privacy in Interactions with Conversational Agents
- Ivoline Ngong
- Swanand Ravindra Kadhe
- et al.
- 2025
- ACL 2025
The ShareLM Collection and Plugin: Contributing Human-Model Chats for the Benefit of the Community
- Shachar Don-Yehiya
- Leshem Choshen
- et al.
- 2025
- ACL 2025