About the Team
The is responsible for various safety work to ensure our best models can be safely deployed to benefit society. It is at the forefront of OpenAI's mission to build and deploy safe AGI, driving our commitment to AI safety and fostering a culture of trust and transparency
As the cutting edge AI models get deployed to the real world at fast speed, we are facing emergent challenges in the security and privacy domains that are specific to large language models.
More research and learning from practical deployment is needed to develop principled solutions for problems, including but not limited to, model inversion or data extraction prevention, knowledge unlearning, anti-regurgitation, fine-tuning safety, privacy-aware data flywheel and protection against data poisoning.
We seek to learn from deployment and distribute the benefits of AI, while ensuring that this powerful tool is used responsibly and safely.
About the Role
We are seeking strong research engineers for pioneering methodologies and implementing systems to reduce risks of various AI security and privacy research challenges during model deployment.
You will have an opportunity to shape the vision of this problem domain, work on the cutting edge of AI research, and collaborate closely with cross-functional teams to improve AI security and privacy protection of our models and systems.
This role is based in San Francisco, CA. We use a hybrid work model of 3 days in the office per week and offer relocation assistance to new employees.
In this role, you will :
Design, implement, and evaluate novel methods to protect AI models and systems from threats such as data extraction and model inversion attacks.
Collaborate with the Post Training team to integrate privacy-preserving techniques into AI model development.
Lead efforts in researching and implementing solutions to mitigate risks proactively associated with data poisoning, membership inference attacks and more.
Work closely with cross-functional teams to establish security and privacy best practices and guidelines for model deployment.
You might thrive in this role if you :
Are strongly motivated by of building safe, universally beneficial AGI and are aligned with
Hold a Ph.D. or other degree in computer science, AI, machine learning, or a related field.
Have 3+ years of experience in the field of AI security and privacy research for deep learning models, especially in areas like membership inference, privacy-preserving ML, adversarial attacks.
Have an in-depth understanding of deep learning research and / or strong engineering skills, particularly proficient in programming languages such as Python and machine learning frameworks like PyTorch (preferred) or TensorFlow.
Stay goal-oriented instead of method-oriented, and are not afraid of tedious but high-value work when needed.
Are a team player who enjoys collaborative work environments.