CapabilitiesSignature practice
Make every claim verifiable: who made it, what it is, and where AI was involved.
How do we show where AI was used, and prove who and what can be trusted?
Synthetic content, autonomous agents and new transparency laws are changing what customers and regulators will accept. We design provenance, identity and disclosure into products and processes, on the open standards the market is adopting.
How AI involvement is recorded in an open content credential. Source: C2PA Technical Specification 2.4; Hkdolts analysis.
The open provenance standard now carries a dedicated, machine-readable assertion for AI involvement: model type, scientific domain and the degree of human oversight. Products that treat disclosure as a badge will have to rebuild it as data.
Researchers have shown that an authentic-looking credential can be produced simply by leaving the AI-origin signal out, with no cryptography broken. The answer is governance: deciding where disclosure is mandatory, and verifying that it is present.
Source: Authenticated Contradictions from Desynchronized Provenance and Watermarking, arXiv, March 2026
Industry guidance now uses the disclosure assertion and per-action AI indicators to tell an original from an AI-assisted edit, going well beyond a binary “AI-generated” flag. Policies written around a single label are already out of date.
Source: C2PA, Use of Content Credentials to Identify Synthetic and Non-Synthetic Content
Five capabilities, each defined by what the client receives.
How AI involvement is disclosed and content origin is proven, designed in line with the open provenance standard.
Obligations from AI and transparency regulation mapped to the controls and metadata a product must carry.
Credentials and vocabularies any party can verify, proven by automated conformance tests, with secure messaging built for a post-quantum future.
Software bills of materials joined to verifiable credentials and trust registries, so suppliers can prove what they ship.
How autonomous AI agents are identified, authorized and paid, on emerging open protocols.
Products whose origin, authorship and AI involvement can be verified by anyone, and a compliance position ahead of the law.
A fixed-fee positioning engagement or a defined diagnostic, then a retainer as standards and regulation move.
Selected work behind this practice.
What we didCore design work on how AI use is disclosed within the open provenance standard, aligned with emerging regulation
The impactAI involvement and content origin become verifiable
What we didBriefings on AI and content-transparency regulation for an industry forum’s working groups, with verified timelines
The impactWorking groups plan against the law as it stands, not as rumored
What we didA multi-year strategy for secure agent messaging, covering post-quantum security and interoperability
The impactGaps are named in the strategy before they reach implementers
What we didCo-researched within an industry working group: portable preference credentials with an executable conformance test suite
The impactHotels, airlines and platforms read one verified preference at a time, with the traveler in control
What we didA phased programme joining software bills of materials to verifiable credentials and trust registries
The impactSuppliers can prove what they ship, and buyers can verify it
What we didResearch and executive briefings on how AI agents are authorized and pay on open protocols
The impactLeaders see how agent-led commerce will be governed
Published perspectives on the questions behind this practice.
A profile of Nvidia’s role in the industrial metaverse: its platforms, the applications they enable, and use cases from digital twins to robotics simulation.
An overview of Meta’s metaverse strategy, its virtual and augmented reality devices, and the social and enterprise platforms built on them.
What Universal Scene Description is, how its object model composes a scene from layers, and the features that let it scale across production pipelines.
Tell us where you stand. We will reply with how we would approach it.
From the United States to leadership teams worldwide, virtually or on site.
