Thinking Through Closed- vs. Open-Weights Model Adoption Impacts on Cybersecurity / Networking

Research evidence excerpt

Thinking Through Closed- vs. Open-Weights Model Adoption Impacts on Cybersecurity / Networking

Update

August 3, 2026 04:01 AM GMT

Morgan Stanley & Co. LLCMCybersecurity / Networking | North America Meta A Marshall

Equity Analyst

Thinking Through Closed- vs. Meta.Marshall@morganstanley.comAntonio Jaramillo +1 212 761-0430

Research Associate

Antonio.Jaramillo@morganstanley.com +1 212 761-4438

Open-Weights Model Adoption Ryan Lountzis

Ryan.Lountzis@morganstanley.com +1 212 761-3189

Impacts on Cyber / Networking Abhishek S Murli

Abhishek.S.Murli@morganstanley.com +1 212 761-7388

Along with our thematic teams note this morning discussing Lucas Cerisola

closed- vs open-weights model adoption scenarios, we lay out ResearchLucas.Cerisola@morganstanley.comAssociate +1 212 761-9194

how we see the scenarios unfolding in terms of cyber/

Software

networking beneficiaries. Enterprise cyber/networking vendors North America

benefit more in open-weights adoption, optical more in closed. Industry View Attractive

Key Takeaways

In general, more traffic / agents / data from AI adoption is a positive for the

networking / cyber ecosystem.

However, believe cloud heavy names more likely to benefit in closed-weights

world, enterprise names more likely to benefit in closed-weights world.

Please see our Global Thematics note today: Weighing In: Open-Weights Models & 3

States of the World.

All AI adoption is good, but believe beneficiaries change based on closed- vs.

open-weight models adopted. AI adoption by enterprises is overall a positive for

both cyber and networking. There are more traffic (networking / network security),

there are more agents (identity security), there are more data centers (more optical

/ networking). This is in addition to just hackers being more enabled with AI