SHARED INTEL Q&A: PKI’s unfinished business—’digital passports’ for content, models and agents

The post SHARED INTEL Q&A: PKI’s unfinished business—’digital passports’ for content, models and agents appeared first on The Last Watchdog.

By Byron V. Acohido
As if keeping track of machine identities wasn’t hard enough.
AI agents are now arriving by the thousands — and most enterprises are just handing them borrowed credentials and hoping for the best.
Meanwhile, the cryptographic infrastructure asked to absorb these threats faces a hard regulatory countdown requiring digital certificates — the credentials securing every machine connection — to rotate eight times more often than they do today.
Related: Certificate expiration is speeding up
That same foundation now faces three converging demands at once: vouch for AI agents, authenticate synthetic media and survive the coming migration to quantum-safe cryptography — all on overlapping deadlines
I sat down with Amit Sinha, CEO of DigiCert, at RSAC 2026 to ask whether the foundation can take it. Sinha runs one of the largest certificate authorities on the planet, which makes him a useful person to put the question to. As it turned out, he was previewing the direction DigiCert would formalize a few weeks later. Edited highlights of our conversation follow.
LW:  Is the machine identity problem actually solved?
Sinha: Not even close. There are roughly 100 times more machine identities than human identities in a typical enterprise. Public TLS certificate validity went from 398 days to 200 in March. By 2029 it’s 47. That’s an eight-fold increase in rotation frequency for a fleet that’s already growing because of data center buildouts. A bank CISO told me they have about three certificate-related outages a day at one-year certs. At 47-day certs, that’s 24 outages a day. You either automate the full lifecycle, or you live with the disruption.
Now layer AI agents on top. Gartner’s view, and ours, is that an agent is essentially a smart workload — it has memory, persistence and it calls APIs. Most enterprises today are letting agents inherit the human user’s identity. That’s a shortcut that doesn’t scale. Every agent needs a digital passport: a cryptographic, immutable identity that travels with it. Once it has that, you can decide what it’s allowed to access, audit what it did, and revoke it instantly if something goes wrong.
LW: What did generative AI change about the trust problem specifically?
Sinha
Sinha: It introduced three new trust problems. The first is content. You can now generate near-perfect synthetic video, audio and images. We’ve been working with Microsoft, Adobe and others on C2PA — the Coalition for Content Provenance and Authenticity. It applies the same PKI cryptography that proves an Apple software update is genuine to a video clip or image: signs the asset, captures who made it and every change, lets you trace it to the source. Where this lands is a world of zero-trust media — any clip is assumed fake unless it carries a signed authenticity record.
LW: The second trust problem is models. What does that look like in practice?
Sinha: Hugging Face alone hosts more than two million models. A developer downloads what looks like a fine-tuned DeepSeek variant and ships it inside the application. Is it really what it claims to be? Has it been tampered with? What’s in the AI bill of materials? What enterprises want is a Lego block of pre-approved, signed models they can compose into their software supply chain and pass governance checks on the way out the door. Without that layer, you’re shipping AI you can’t fully account for, and regulators are going to start asking pointed questions.
LW: And the third — agents. What does extending cryptographic identity to autonomous agents actually look like?
Sinha: Standards the PKI community has been refining for years — SPIFFE and SPIRE, frameworks for issuing identities to dynamic workloads at high volume — extend naturally to agents. You treat the agent as a unit, give it a cryptographic identity, then layer authorization and enforcement on top. Agent-to-agent protocols are starting to support agent passports, where one agent cryptographically proves it is the ServiceNow agent or the Salesforce agent it claims to be. What’s missing today is that durable foundational identity. Most deployments are leaning on bearer tokens borrowed from the human’s SSO session. As one CISO put it: we spent years training humans not to click on things, and now there’s an AI agent clicking on things 100 times a second.
LW: Does AI change how much time the industry has to absorb all this?
Sinha: It’s a wake-up call. What AI has done is time dilation — what used to take a year happens in weeks. Meanwhile the foundational pieces have been relatively static. They’re like the plumbing in your house: you don’t redo it every time you remodel a kitchen, but every once in a while you have to look at it seriously. The 47-day TLS mandate lands in 2029. The migration to quantum-safe cryptography lands in roughly the same window. On top of that, organizations are building trust frameworks for content, models and agents all at once. We’re in a once-in-30-year upgrade cycle. The encouraging part is that I no longer have to explain why any of this matters to CISOs or boards. They’re already asking. The question now is execution.
* * *
PKI made e-commerce viable when it was still a novelty, then carried the internet through cloud and mobile. The cycle Sinha describes asks the same foundations to absorb three new categories of trust simultaneously — cryptographic provenance for media, signed integrity for models, durable identity for agents. Boards are in the room now. The question is execution.
I’ll keep watch and keep reporting.

Acohido
Pulitzer Prize-winning business journalist Byron V. Acohido is dedicated to fostering public awareness about how to make the Internet as private and secure as it ought to be.
(Editor’s note: I used Claude and ChatGPT to assist with research compilation, source discovery, and early draft structuring. All interviews, analysis, fact-checking, and final writing are my own. I remain responsible for every claim and conclusion.)

April 30th, 2026 | Q & A | Top Stories

*** This is a Security Bloggers Network syndicated blog from The Last Watchdog authored by bacohido. Read the original post at: https://www.lastwatchdog.com/shared-intel-qa-pkis-unfinished-business-digital-passports-for-content-models-and-agents/

About Author

AndyC

Andy Curtis is an award-winning security consultant, researcher and public speaker. He has been working in the computer security industry since the early 1990s, having been employed by state and federal government, leading healthcare and banking providers across three continents. He has given talks about computer security for some of the world’s largest companies, worked with law enforcement agencies on investigations into hacking groups, and is a regular voice on TV and radio explaining IT security threats.

See author's posts