Ben Harack

Ben Harack

I study the potential for artificial intelligence to trigger a world war and how to prevent that from happening.

Most recently, I have led a year-long investigation into AI verification, where my team and I found that AI verification is neither impossible nor trivial. Workable systems can be built within a few years if there is political will toward this goal. Moreover, we found that AI verification can drastically open up the space for AI – both for governance and for the industry itself. The opportunities of this set of technologies are manifold, and I hope that countries and companies can be convinced to pursue them.

My background includes natural, formal, and social sciences, with a particular emphasis on machine learning, semiconductors, cryptography, and nuclear science. Previously, I also spent a decade working for Silicon Valley startups as a software engineer and manager. Presently I am completing my DPhil in International Relations at Oxford University.