You can't secure a corporate black box you aren't allowed to see. Open source lets thousands of devs audit the code and fix flaws instantly, making AI progress democratic and safe.
Open source also reduces dependency on a few gatekeepers. If a model breaks, biases users, or gets worse, communities can fork it instead of waiting on a vendor.
