Recent advances in AI surprised many observers. We expect more surprises to come, and we think society needs to do more to prepare for them.

As AI models advance in capability, we expect their impact on society to grow. Managed well, such advances can offer benefits in fields as diverse as education, medicine, and clean energy. However, they could also pose serious risks by failing to perform as expected in critical applications, or by helping malicious actors launch cyberattacks, spread disinformation, or develop biological weapons.

We think that reducing the risks of advanced AI is one of the most important questions of our time. We also think it is a highly interesting problem, with many opportunities for people to make progress on it through technical work, policy research, and policymaking.

The Georgetown Responsible AI Network is keen to support individuals who want to work in this field.