Image credit: Pexels
As deepfakes, synthetic identities, and AI-powered fraud become more sophisticated, five innovators are advancing cryptographic verification, behavioral biometrics, and layered authentication to strengthen digital trust.
As artificial intelligence makes deepfakes, synthetic identities, and voice cloning increasingly convincing, verifying identity online has become one of the most pressing challenges for businesses, governments, and digital platforms. Traditional methods built around document inspection, passwords, and static credentials are proving insufficient against adversarial AI. In response, a new generation of identity innovators is developing systems that rely on cryptographic verification, behavioral biometrics, layered authentication, and open standards to establish trust without sacrificing privacy.
Together, five companies and technology leaders are advancing approaches that move beyond simply identifying users. Their solutions focus on creating verifiable, portable, and auditable trust for an internet where both humans and AI agents increasingly interact.
Replacing Document Inspection With Cryptographic Proof
For years, identity verification has depended largely on visual inspection of government-issued documents. That approach is becoming more vulnerable as AI-generated forgeries grow more sophisticated.
Aho is addressing this challenge through cryptographic proof of origin using Verifiable Credentials (VCs) and C2PA standards. Rather than requiring users to upload complete identity documents, the company’s selective disclosure capability enables platforms to request only the information they actually need. For example, a service can verify whether someone is over 21 without revealing their full date of birth or other personal information.
“The current way they’re doing it is document inspection. The better way to do it is through cryptographic verification,” said Scott Goci, Founder and CTO, Aho.
He added, “With selective disclosure, you only have to reveal the exact information that is being requested by the site and nothing else. They can ask your Apple Wallet, ‘Is this user over 21?’ and all they get back is yes or no.”
Looking beyond today’s verification challenges, Aho is also preparing for AI-driven digital identities.
“Cryptographic verification (until quantum computing kind of breaks it) should be the gold standard. It’s all about building a composable chain of trust where verified credentials from multiple trusted sources create a reliable digital ecosystem.”
With universities already using the platform to verify academic transcripts, Aho is also developing a roadmap for agentic identity to support trusted AI accounts.
Making AI Decisions Auditable
As AI becomes more involved in hiring, lending, and business decision-making, understanding how those decisions are made is becoming equally important.
The Machine Experience (MX) Community has developed the Community-Owned Governance System (COGS), which embeds a complete human-readable audit trail directly into documents. Every AI interaction, including prompts, responses, and decisions, is recorded. Its accompanying tool, Reginald, cryptographically seals that information, allowing documents to be independently verified even years later.
“If you ask an AI why it made a mistake, it doesn’t tell you why it made the mistake — it invents a reason, because AIs are prediction machines, not really thinking machines,” said Tom Cranstoun, Founder of Machine Experience (MX) Community.
He explained, “The document contains the audit trail in metadata — who created it, when, what AI was involved, what prompts and responses were sent. Even if my company were to disappear tomorrow, the document contains the information, and it’s an open standard that anybody can read.”
Cranstoun also argued that the industry needs common standards.
“At the moment, if you’re working with Claude, a different set of rules from OpenAI, a different set of rules from Google — everybody’s doing their own thing. If we have open standards and a model to follow, then we improve it for both the humans and the machines.”
Combating AI-powered Voice Scams
AI voice cloning has transformed fraud from isolated scams into highly scalable attacks capable of targeting thousands of victims simultaneously.
Trust Onion‘s response combines rotating three-word authentication codes generated locally on users’ devices with “Proofies,” biometrically verified selfies containing embedded time, date, and location information. Designed with simplicity in mind, the application does not require account creation to access verification codes, reducing barriers for elderly and vulnerable users.
“Scams in the past were one person trying to scam another — one to one. With an automated AI bot, in theory, there can be a thousand phone calls happening at once. It’s just a numbers game where the threat level is higher and the quantity is higher,” said Jesse Seaver, Co-Founder and CEO, Trust Onion.
He emphasized the privacy benefits of local authentication.
“Those words are not calculated on the server or stored on the server. It’s just calculated locally on your phone, and it even works in airplane mode. It can’t be intercepted.”
Seaver also highlighted the company’s focus on accessibility.
“The app is designed to be very low friction. You don’t have to create an account to see the code words. We remove those barriers by design so that older people in particular have a higher success rate of getting to that code word.”
Beyond consumers, the platform also offers white-label deployments for enterprise teams and insurance providers seeking to deliver identity protection services.
Accelerating Trusted Verification Globally
While identifying fraudulent documents remains important, Trinsic focuses on making legitimate identity verification faster, reusable, and internationally scalable.
The company aggregates cryptographically verified digital identities issued by governments and trusted institutions through a single API. Instead of relying on document scans that can be manipulated, organizations receive digitally signed credentials that can be verified in as little as 15 seconds.
“Traditional verification methods — document scanning, selfie uploads — are slow, failure-prone, and create significant user friction. They’re also completely vulnerable to AI-generated deepfakes and synthetic identities,” said Benjamin Cejvan, Communications and Marketing Professional, Trinsic.
He added, “Our credentials are cryptographically proven by governments and trusted organizations. AI cannot forge a credential that’s been digitally signed by a government. We can verify in as little as 15 seconds, and that credential cannot be faked.”
As digital identity regulations expand across Europe, Trinsic is positioning itself to simplify cross-border compliance.
“The EU’s AMLR and EUDI initiative are going to mandate stricter identity checks. We’re prioritizing integration with all 28 EU nations’ digital wallets as they launch — one API, one contract, and a dashboard configuration instead of months of legal and engineering work for each country.”
Building Continuous Trust Through Behavioral Intelligence
Thomas Prommer argues that identity verification should not end after an initial login. Instead, trust should evolve continuously according to user behavior and changing levels of risk.
His approach combines traditional biometric verification with behavioral biometrics, analyzing characteristics such as phone angle, grip patterns, and typing cadence to establish a multidimensional trust profile that becomes increasingly difficult for AI-generated impersonations to replicate.
“Static verification is no longer secure. Sending passport screenshots is unsafe. Modern verification requires dynamic, three-dimensional interactions — face movement, blinking — to prove liveness and prevent spoofing,” said Thomas Prommer, AI executive and Group CEO of We The Flywheel.
He explained, “Behavioral biometrics create a multi-channel, multi-dimensional trust profile. AI analyzes unique patterns from how you interact with your device — phone angle, grip, typing cadence — to trigger additional validation when something seems off.”
Looking ahead, Prommer believes organizations will increasingly collaborate to reduce repeated verification while maintaining security.
“Companies will form alliances to create shared, trusted identity registries. Users complete verification once with a trusted party — like ID.me — and can then use that validated identity across many apps without repeated friction.”
Trust Becomes the New Digital Infrastructure
Static, document-based identity verification is no longer sufficient in an era of AI-generated deception. Cryptographic proof, behavioral signals, privacy-preserving credentials, open standards, and continuous authentication are rapidly becoming the foundation of digital trust.
While each innovator approaches the challenge differently, they share a common objective: building an internet where trust is verifiable, portable, and resilient enough to keep pace with the growing, sophisticated AI-driven threats.
