Sep 22, 2025 - Axios Events
Axios AI+ DC Summit: Specificity is key to successfully managing AI risk, experts say
Add Axios as your preferred source to
see more of our stories on Google.
Attendees gather around the table for the discussion. Photo: Bryan Dozier on behalf of Axios.
WASHINGTON, D.C. – Experts at Axios' AI+ DC Summit agreed that a major key to managing AI risks is to tailor solutions around individual settings and situations and avoid the one-size-fits-all mindset.
- Axios' Megan Morrone and Ashley Gold moderated the roundtable discussion in Washington, D.C., on Sept. 17. The event was sponsored by the Center for Audit Quality.
Why it matters: AI only works in an organization if everyone can trust the results — and that takes clear rules.
5 big things: Here's what the attendees had to say …
- Agentic AI is changing how companies think about managing the risks and sensitive data, UserTesting VP and head of AI Michael Domanic said. With agentic AI, "you're increasing the surface area of potential risk, but you're also exponentially increasing the potential value that you might receive by doing these things."
- Many issues surrounding AI risk are still unresolved, said EqualAI president and CEO Miriam Vogel. "I think the problem is we're continuing to see new challenges to us personally, to our work," she said. "We're certainly seeing the benefits, but we're not answering the need, the concerns on so many of these risks.".
- One of the biggest challenges is that risk appetites are so large, said Miranda Bogen, Center for Democracy & Technology AI Governance Lab director. "We often worry that risks end up being prioritized in ways that aren't actually addressing the concerns and the harms that are impacting people who will, I think, bolster this market-level trust that we want, so we can't forget that dimension."
- Risk management is going to vary depending on the situation and industry, Business Software Alliance managing director Kate Goodloe pointed out. "It makes sense when we think about what risks are involved and what mitigation strategies do we apply to support trust, that they're going to look very different in these different scenarios."
- Identifying risk is very domain-specific and requires distinct knowledge, Center for Democracy & Technology vice president of policy Samir Jain said. "When we talk about risk, we shouldn't think about it just as risk," he said. "It's really a very specific issue that actually requires a whole lot of expertise."
Sponsored content:
Center for Audit Quality CEO Julie Bell Lindsay highlighted how important trust is to AI acceptance.
- "Our capital markets in the U.S. [and] globally, they depend on trust. … I also know that one of the inhibitors to adoption and deployment of AI is lack of trust," Lindsay said.
