Platform partners: adding a physical layer to content authentication
Daniel Oblitas Garafulic · · Updated · 4 min read

In short: detection software can only estimate whether content is fake. Lightmark adds a signal from the physical world: a dynamic optical signature that is captured with the scene. Partners integrate it into what they already sell.
The problem at scale
AI-driven impersonation and synthetic-media fraud are now a measurable cost for platforms, brands and institutions. Software-only detection is a classifier problem: as generators improve, accuracy on real-world content falls, and every new model resets the race.
Regulation is raising the stakes. Since 2 August 2026, Article 50 of the EU AI Act requires AI-generated and manipulated content to be marked and disclosed. Buyers of detection and trust products are now asking a sharper question: what evidence supports this verdict?
How the integration works
- Embed. Optical emitter modules integrate into the partner's hardware, lighting or camera-side stack. No bespoke camera firmware is required.
- Capture. Standard CMOS sensors record the dynamic signature together with the scene.
- Verify. The partner's platform calls Lightmark's verification service, which reads the signature per pixel and returns provenance and integrity results.
Three integration partner patterns
| Partner type | What they integrate | Commercial model |
|---|---|---|
| AI detection and authentication platforms | Lightmark verification as a ground-truth signal inside their product | Licensed per deployed customer environment, with joint go-to-market |
| Device, camera and lighting OEMs | Emitter modules built in at manufacture | Licensed per device shipped |
| Media, broadcast and government platforms | Newsroom, broadcast and compliance workflows | Enterprise pilot-to-production terms |
All three produce the same output: a signed result showing whether content originated in an authenticated optical environment.
Where Lightmark sits next to the alternatives
| Lightmark | Camera-embedded signing | C2PA content credentials | Watermarking | |
|---|---|---|---|---|
| Where it operates | Physical layer, before capture | In the device | File metadata | File pixels, after capture |
| Works with existing camera fleets | Yes | New cameras only | Depends on tooling | Yes |
| Ground truth for classifiers | Designed to provide it | Limited | No | No |
| Survives re-encoding and re-upload | Designed to (validated in pilots) | Metadata often stripped | Metadata often stripped | Variable |
| On-premises deployment | Yes | Varies | Varies | Varies |
C2PA, watermarking, camera signing and deepfake detection each solve part of the problem. None of them operate before capture, on the cameras organisations already own.
Independent analysis of the C2PA gap
Recent academic security analyses of C2PA have argued that content credentials alone do not meet their security goals for high-stakes uses, and can mislead users and policymakers if relied on prematurely. Lightmark is complementary to C2PA: credentials document a file's history, Lightmark provides physical evidence of capture.
What integration partners get
- A differentiating evidence layer competitors cannot replicate in software
- Named partner status and co-marketing
- First-integration privileges in their category
- Revenue through per-device, per-customer or per-deployment licensing
Working with Lightmark
- Partnership discovery — a 30-minute briefing on architecture and commercial structure.
- Scoped pilot — integration into one product line or customer environment, with technical support and a verification report.
- Programmatic partnership — organisation-wide integration, co-marketing and licensing or revenue share.
Frequently asked questions
What is an integration partner?
A company that embeds Lightmark's emitter or verification capability into a product it already sells, and licenses it to its own customers.
Do partners need to change their product stack?
No. Lightmark is designed to sit alongside existing detection, capture and distribution pipelines.
How is Lightmark different from C2PA, watermarking and camera-embedded signing?
Those methods operate on the device or the file. Lightmark operates on the light in the scene, so it works with existing cameras and does not depend on metadata surviving distribution.
How is the commercial model structured?
Per customer environment for platforms, per device for OEMs, and enterprise terms for media, broadcast and government platforms.
Ready to explore a partnership? Apply to the partner programme or book a briefing.
See it on your footage.
Book a briefing
