Sarah de Haas is a Product Manager at Google Research, based in London. Her work focuses on creating verifiably private, secure, and safe Artificial Intelligence systems from the hardware design through to the interaction with end users. She is a lead on Google’s Project Oak, an open-source system that provides TEE-based (Trusted Execution Environment) runtimes that allow for externally inspectable data usage and flow in a verifiable way. Previously, she worked on counter abuse systems at Google, and on AI productionisation at DeepMind. She is particularly interested in technologies that balances privacy with monitoring systems for safety, and the socio-technical systems that make use of them. In her spare time, she also lends her product and program management expertise to Project CETI, an international effort across multiple university research groups using cutting edge AI techniques to understand the language of sperm whales.