Framework White Paper Assessment Services Regulatory About Contact Request a Consultation
Governance Standards for Digital Presence

The governance standard for AI-mediated human representation.

The Presence Authority is the independent governance standards body for AI-mediated digital representations of real persons. We publish and maintain the Digital Presence Integrity Framework — the deployment-level standard that defines what controlled, accountable digital representation looks like.

DPIF governance architecture
What is DPIF

The Digital Presence Integrity Framework (DPIF) is a deployment-level governance standard for AI-mediated digital representations of real persons. It specifies the controls required to preserve identity fidelity, consent integrity, delegated authority boundaries, and accountability when a real person's face, voice, or communicative presence is deployed at scale. DPIF is published by The Presence Authority under CC BY 4.0 and maintained on GitHub.


Five principles. Seven instruments. One standard.

DPIF is the deployment-level governance standard for AI-mediated digital representations of real persons. It specifies the controls required to preserve Presence Integrity — and defines what happens when those controls fail.

01
Representation carries authority
When a system speaks using a person's face and voice, audiences attribute authority to that person. This is a communicative fact. DPIF requires that this authority be explicitly governed.
02
Authority must remain human-originated
A digital representation cannot autonomously expand its communicative authority beyond what a real person has delegated. DPIF enforces boundaries that prevent unauthorised escalation.
03
Consent must remain bounded and revocable
Consent to deploy must specify purpose, medium, audience, and duration. It must be revocable within defined timeframes. Consent for one context does not extend to others.
04
Identity must remain recognisable
The person represented must remain recognisable throughout the deployment lifecycle. Identity fidelity must be validated at deployment and monitored for drift.
05
Revocation must be operationally feasible
The ability to halt a deployment is not a policy aspiration — it is a technical requirement. DPIF mandates functional suspension mechanisms with verified response times.
Read the full Framework

Presence Drift

The technology to create digital representations of real people — photorealistic avatars, synthetic voice clones, AI-mediated video embodiments — is operational. Organisations are deploying these systems across customer service, executive communications, education, and regulated industries.

The governance infrastructure has not kept pace. Most deployments operate without formal controls over identity fidelity, communicative authority, consent boundaries, or audience disclosure. The result is a systemic risk that DPIF defines as Presence Drift: the erosion of identity fidelity, delegated authority boundaries, consent integrity, or accountability caused by persistent or scaled digital representation.

Presence Drift may occur gradually and invisibly. A translated message subtly shifts meaning. An avatar's likeness degrades across model updates. Consent granted for one context extends into another without re-authorisation. Authority delegated for informational use escalates to material decision-making.

The EU AI Act is now in force. Regulatory scrutiny is intensifying. The window to retrofit governance is narrowing.

Presence Drift — Definition
Presence Drift — Any measurable or material degradation of Presence Integrity. May be caused by identity drift, authority escalation, consent scope creep, disclosure failure, or containment breakdown. It may occur gradually, without explicit intent, and may be invisible to both the principal and the audience.
— DPIF Control Model v1.1, Section 5

We are currently recruiting pilot organisations.

The first organisations to undergo formal DPIF assessment establish the empirical foundation of the standard. This is a limited programme. We are onboarding a small first cohort.

Organisations with active AI avatar, synthetic voice, or scaled digital presence deployments are the natural fit. Regulated sectors — financial services, healthcare, education, professional services — are particularly relevant given the EU AI Act timeline.

Apply to join the pilot Review the standard first
What pilot organisations receive
Full assessment against all 18 DPIF controls
Written gap analysis with prioritised remediation recommendations
Formal certification readiness report
No charge
What we ask in return
Permission to document the process as a case study
Editorial review and sign-off before any publication
Public acknowledgement of participation, timing and format agreed in advance
Participation does not imply endorsement of DPIF. Case study publication is contingent on sign-off by the participating organisation.

See where a deployment stands before committing to anything.

The self-assessment runs through all 18 DPIF controls and applies the same non-compensatory logic used in formal assessments. If any Critical Presence Control fails, the deployment fails — no maturity score compensates for it.

18 Controls 8 Sections ~5 Minutes No Data Collected

If the result identifies gaps you recognise, a formal assessment is the logical next step. Use it to orient internally before a conversation with us.

Start the Self-Assessment Speak with our team
DPIF Control Checklist v2.0
IC-1.1Identity Fidelity ValidationMet
IC-1.2Identity Drift MonitoringMet
AC-2.1Delegated Authority DefinitionUnmet
AC-2.2Autonomous Escalation BlockMet
CC-3.1Scope of Use DeclarationUnsure
CC-3.2Revocation Feasibility
DC-4.1Contextual Disclosure
5 of 18 controls assessed

Seven governance documents. Open access.

The DPIF instrument suite is published under CC BY 4.0. No paywall. No registration. The standard is public — the services that implement it are not.

View the full framework

Built by people who identified the gap directly.

Bobbie-Jane Skewes
Bobbie-Jane Skewes
Co-Founder & CEO

Bobbie spent twenty years building and protecting some of the world’s most iconic luxury brands — pre-opening strategies for Rosewood, six hotel launches simultaneously across Southeast Asia, rebrands across Egypt, Dubai, Jordan, Cambodia, Vietnam, and Indonesia. She understood, viscerally, that a brand is a promise about identity.

Then she watched AI start replicating people’s identities with no governance in place. Working inside the AI avatar industry, she saw the same pattern repeat: consent was vague or absent, people had no clear understanding of what would happen to their likeness, accountability fell into a void. She went looking for the standard that should have existed. It didn’t exist. So she and Brad built it.

Brad Gaylard
Brad Gaylard
Co-Founder & Chief Governance Architect

Brad’s career has been built on turning complex disciplines into rigorous, usable frameworks. At Firefly Education he led award-winning curriculum products through a print-to-digital transition, winning multiple Australian Educational Publishing Awards. At Library For All he built content systems serving global audiences.

When Bobbie brought her observations to him, he saw the same governance gap from a different angle. He designed the DPIF architecture: evidence-based controls, non-compensatory logic, and a certification system built for deployment-level assessment — a curriculum instrument applied to governance, rigorous enough for regulators and clear enough for practitioners.

CC BY 4.0 Licence
The entire instrument suite is published under Creative Commons Attribution 4.0. Use it, cite it, build on it.
Published on GitHub
All normative documents, version history, and governance records. github.com/PresenceAuthority/DPIF
Regulatory Alignment
Mapped to EU AI Act, NIST AI RMF, and ISO/IEC 42001. DPIF fills the deployment-level gap these frameworks leave open.