09/09/2026
Dr. Carina Prunkl, Lead Writer of the 2026 International AI Safety Report, presented key findings from the report during the International AI Safety Convening at Daystar University.
She highlighted the rapid and uneven improvement of AI systems, including advances in mathematics and science and the increasingly realistic generation of images, audio and other content.
The briefing also examined emerging challenges, including the use of AI-generated content for impersonation, fraud and harassment, as well as the growing usefulness of AI in cybercrime.
Importantly, Dr. Prunkl clarified that the report presents findings rather than recommendations. Its purpose is to provide an evidence base for understanding the capabilities, risks and potential safeguards surrounding increasingly advanced AI systems.
She further noted that societal resilience can play an important role in preventing and mitigating harm as AI risks move from theoretical concerns to practical challenges.