Faculty at @georgetownsfs and Senior Fellow at @CSETgeorgetown. Author of three books on cybersecurity, AI, and national security. Personal account.
May 19, 2021 • 11 tweets • 4 min read
Last year @CSETGeorgetown approached @OpenAI with a question: could we see whether GPT-3, their prized AI writing system, would generate disinformation? To my great surprise, they said yes.
Our report today shows how GPT-3 can spin, distort, and deceive.
cset.georgetown.edu/publication/tr…
First, we tested if GPT-3 could write tweet-length messages arguing that climate change was a hoax. The answer was unambiguously yes. GPT-3 effortlessly wrote about the “climate mafia,” compared climate science to communism, and said that wind power kills.
Dec 18, 2020 • 6 tweets • 2 min read
I’m hesitant to do this because everyone says how the SolarWinds hack confirms all their priors, but here goes.
My PhD and first book argues the blurred lines between espionage and preparations to attack in cyber make interpretation hard. A short thread on what that means here.
Simplifying a lot, the problem is this: the illicit network access hackers use to spy can also enable attacks. Not particularly high-end ones, but still damaging.
This means that it’s sometimes hard to know the intentions behind an intrusion. Was it espionage or a foiled attack?
