Innovation rarely fails all at once. More often, it narrows: fewer organizations can afford to participate, fewer technical paths remain available, and decisions that once belonged to builders move behind legal, financial or proprietary walls.
That concern sits at the center of Jonathon Marshall’s essay, The AI Regulatory Moat: When Innovation Stalls, Lawyers Take Over. Marshall argues that regulation designed around inputs—who may build, how much compute they may use, and which compliance structures they can afford—can protect established companies while making meaningful experimentation harder for smaller teams and open-source communities.
A warning about closed paths
Marshall’s argument is not that safety, accountability or oversight should disappear. It is that the structure of those safeguards matters. Rules focused on measurable harm and transparent outcomes can protect people. Rules that make participation unaffordable can turn safety into a barrier that only incumbents can cross.
“Innovation cannot be regulated into existence. But it can be regulated into extinction.”
Jonathon Marshall
The distinction is especially important in artificial intelligence, where a small number of organizations already control enormous concentrations of compute, data and capital. If the next generation of research must also pass through frameworks designed by those same organizations, technical stability may come at the cost of technical possibility.
The exchange that started this conversation
Marshall shared the article directly with Brian Davis, President of Aligned Technology Partners. Davis’s immediate response was candid:
“Thats pretty deep man”
Brian Davis - President of Aligned Technology Partners
The short reply reflects a larger point. The argument reaches beyond AI policy. Every technology market eventually faces pressure to exchange openness for control and experimentation for predictability. Stability matters—but lasting stability cannot depend on one vendor, one architecture or one organization retaining permanent authority over the future.

Why open source strengthens stability
Open source is sometimes described as the opposite of control. In practice, well-governed open source distributes control. It gives organizations the ability to inspect what they depend on, test claims for themselves, identify failure modes and continue operating when a commercial relationship changes.
That creates several forms of stability:
- Technical stability: code and standards can be inspected, tested and improved by more than one organization.
- Operational stability: teams can document, reproduce and repair systems without waiting for a single vendor.
- Economic stability: competition remains possible because knowledge and interoperability are not entirely enclosed.
- Strategic stability: organizations retain an exit path when a platform, license or provider no longer serves them.
None of this means that every open project is secure, mature or appropriate for every workload. Openness is not a substitute for engineering discipline. It is the condition that makes independent verification, shared maintenance and informed choice possible.
Growth requires more than one permitted answer
The technology industry advances when different ideas can be attempted, compared and discarded. Open-source communities make that process visible. They give researchers and builders a place to challenge dominant assumptions, explore alternative architectures and contribute improvements that are not limited to a single product roadmap.
“True safety comes from diversity, scrutiny, and competition, not from centralized control.”
Jonathon Marshall
For OSIT, this is why open architecture is a business principle rather than a preference for a particular license. The goal is not to remove governance. The goal is to ensure governance does not eliminate portability, transparency or the next good idea before it has a chance to prove itself.
Keep the field open
Responsible technology growth needs safeguards. It also needs room for independent builders, smaller companies, researchers and customers to participate. The strongest systems are not stable because change is impossible. They are stable because change can happen without starting over—or asking one gatekeeper for permission.
That is the practical promise of open source: not freedom from responsibility, but freedom to inspect, improve and choose.
This OSIT perspective responds to Jonathon Marshall’s independently published essay.
Read the original article on LinkedIn →