Five Steps for Secure Adoption of Agentic AI

On July 30, 2026, partner Matthew F. Ferraro published an article with Dow Jones Risk Journal on “Five Steps for Secure Adoption of Agentic AI.” In the article, Mr. Ferraro – who counsels clients widely on the responsible adoption and deployment of artificial intelligence – explained that a compromised AI agent can do substantial damage before any human detects it.

He also discussed five core actions companies should take when considering deploying AI agents:

1. Evaluate the security risks;
2. Update AI frameworks;
3. Implement security best practices across AI agent life cycle;
4. Adopt advisedly and govern aggressively; and
5. Monitor the evolving regulatory landscape.

To read the article, click here. For questions around mitigating the risks of AI while leveraging their benefits, please contact Mr. Ferraro.