Open source AI backers launch $10 million ad campaign
Add Axios as your preferred source to
see more of our stories on Google.
/2025/02/21/1740101218822.gif?w=3840)
Illustration: Aïda Amer/Axios
The Open Source AI Foundation has launched a $10 million ad campaign aimed at convincing policymakers and others of open source AI's benefits, Axios has learned.
Why it matters: There is a spirited debate in both technology and policy circles as to whether open source AI makes the technology safer or less secure.
State of play: The foundation, a new effort led by Cambridge Analytica whistleblower Brittany Kaiser, aims to convince lawmakers that civilian agencies should abandon work with closed source AI companies and adopt open source technologies instead.
- Asked whether the foundation had received backing from tech giants, Kaiser told Axios in a statement, "Not currently — though some of the big five have already expressed interest, and we invite anyone who shares our mission to join us."
What they're saying: "All closed-source AI contracts with civilian agencies should be terminated immediately. Government AI should be built openly, with transparency and auditability at its core," Joe Merrill, CEO of OpenTeams, said in a statement. "This allows the models and training to be publicly scrutinized and verified."
- "Large language models are built on open frameworks and are trained on public data from the web," Eliza Labs CEO Shaw Walters said in a statement. "Closed models are fine for other use cases. But not for public service."
- "Open-source AI allows the public to audit and verify algorithms, enhancing trust in government technology," Quansight CEO Travis Oliphant said in a statement. "It is also more secure because any attacks or exploits can be identified and remediated, while the models and their training can be audited to minimize bias."
The other side: Critics of open source AI argue it is inherently hard to limit and control and will enable bad actors and foreign rivals of the U.S. to evade any limits or protections.
