TC Niedzialkowski’s Post

Hi everyone! I wanted to share a readout from UC Berkeley's Center for Long-Term Cybersecurity and Berkeley Risk and Security Lab event "AI Enabled Cybercrime" sponsored by Fortinet. The full day event included representatives from local government, CISA, FBI internet crimes division, academia, and private industry. There was a short briefing on the current state of AI enabled cyber threats and where experts see it going, plus a questions and answers period. The majority of the event was two sets of tabletop exercises: designing and executing an attack as an APT, and then responding to such an attack on behalf of the targeted organization. In regards to "AI enabled cybercrime" threats the general message was, at least as of today, apart from deepfakes and prompt injection the impact of AI enabled cybercrime is "lowering the bar" for execution e.g. less misspellings in phishing emails, more localized/personalized phishing emails, and attacker/researcher assistance in vulnerability research or configuring/managing hacker tools. The future impact of AI is "yet to be seen" e.g. agentic AI agents or language model breakthroughs. The most valuable part, to me, was networking and learning from diverse perspectives. The tabletop format worked great because I got to see everyone's unique experiences and perspectives in action. I highly recommend engaging with programs such as Berkeley's Center for Long-Term Cybersecurity and Berkeley Risk and Security Lab to expand you and your organizations network and resources!  PS - they are looking for partners and funding!