An artificial intelligence security firm spun out of Carnegie Mellon University and co-founded by an OpenAI board member is expanding its footprint and workforce in Pittsburgh.
Gray Swan AI is subleasing more than 19,000 square feet in Bakery Square along “AI Avenue” in a deal closed by real estate firm Jones Lang LaSalle. The East End AI company plans to “at least double” its Pittsburgh team of more than 25 by the end of the year, fueled by a $40 million Series A funding round in May, according to a Thursday announcement.
“Strong AI security work is happening in Pittsburgh, and that comes down to the people,” said Matt Fredrikson, CEO and co-founder of Gray Swan. “Our work depends on the local AI research talent, so we’re committed to the city where we started.”
The move quadruples Gray Swan’s footprint from its former 4,200-square-foot Shadyside office.
Andrew Millberg and Geoff Greco, senior vice presidents at JLL, contracted Gray Swan’s sublease with terms that allow for expansion as Gray Swan hires. Millberg said the transaction is exemplary of the kind of office flexibility local startups need.
“Fast-growing companies like Gray Swan often have to choose between the space they need today and the space they’ll need in two years,” Millberg said. “We were able to find and negotiate an opportunity that offers them both. They’ll be able to scale incrementally as needed and never have to sacrifice momentum for real estate reasons.”
Read More
CMU professors Fredrikson and Zico Kolter founded Gray Swan in 2024, two years after OpenAI’s ChatGPT ruptured the status quo. Kolter is also director of CMU’s Machine Learning Department and sits on the board of OpenAI as chair of its safety and security committee.
“I have a whole lot of concerns,” Kolter told the Post-Gazette in an In Conversation With Q&A last year. “My concerns are what guide a lot of the research I do, and a lot of the day-to-day work that I pour myself into.
“With this new technology you have this incredibly powerful tool at the fingertips of everyone. Of course, there are going to be massive risks to it and massive safety concerns to it. And I spend a lot of my research and time focusing on basically how we can deploy AI in a more safe and secure fashion.”





Comments
Welcome to Post-Gazette subscriber-only commenting community. Please review our community guidelines.