top of page

Cracking the Code of AI Security: A Conversation with Harriet Farlow


1) Give a brief introduction about yourself, and what inspired you to create Mileva.

Sometimes I introduce myself by saying I missed the boat on computer hacking, so now I hack AI instead. This is a bit tongue and cheek because for eight years my career has been at the intersection of data, cyber, and AI, but I always felt like I’d just missed the computer hacker wave. It was only when I started my PhD that I became aware of adversarial machine learning and AI security and realised I’d found my purpose. AI security refers to the technical and governance practices that secure AI systems from attacks by an adversary, and adversarial machine learning is the discipline that researches those attacks (and subsequent defences). My career has spanned consulting, academia, a start-up and the Australian Government. I have a Bachelor of Science in Physics and Bio-anthropology, a Master's in Cyber Security, Strategy, and Governance, and I’m completing my PhD in Machine Learning Security. I’m currently based in Canberra but have also worked in Sydney and in the USA, - the Bay Area and New York City. I see a clear parallel between the rise of the internet and subsequent cyber security threats and the rapid adoption of AI now, and I want to prevent AI security from becoming the next great threat. I spent a lot of my early career trying to figure out why, in all of my roles, I always felt a bit like a square peg in a round hole. Then I found AI security, realised this is the only thing I want to work on, and decided to go out and make my own box.

2) What have been some of the challenges you have come across on this startup journey?

I’m really privileged and grateful that almost everyone I’ve told about this venture has been very supportive (or has at least done a good job of pretending to be supportive). I know that other founders can struggle to get support from their friends and family, so I’m immensely grateful that I have a circle of incredibly supportive advocates for Mileva. (Moral support may be underrated but it's very important!) Something we did struggle with early in our journey was being taken seriously - some people were skeptical that we were doing something worthwhile, or that we were the right people to do that. I found that these points of view were worth listening to and understanding, but at the end of the day, it’s essential to figure out what advice is worth listening to and what advice we should ignore.

3) Congratulations on winning the New Wave Showcase and Innovation Award! What will be the next steps to help Mileva grow?

Thank you! We’re so humbled to have won two prizes, we were not expecting it at all, and we are so grateful to the donors of the prizes. We have some exciting things in the works that I can’t announce yet, but otherwise, we’re focusing on growing our customer base and continuing to learn about their priorities and pain points. We have ambitious plans for Mileva.

4) For someone who is thinking about creating their own startup, what advice do you have for them?

Just do it! My favourite quote is - ’shoot for the stars, and if you miss, you’ll land on the moon’*. Aim high and work your hardest, and if for some reason it doesn’t work as you imagine, you’ll still be a lot closer than if you hadn’t tried at all.

*the original quote is the other way around but scientifically inaccurate so I rephrase it as this

26 views0 comments


bottom of page