The $27.2 billion problem: Why AI is reshaping digital identity and trust

AI is altering the way work gets done — rapidly, profoundly, and permanently. Processes that once took weeks of manual effort now take hours thanks to AI-driven automation.

At Lumin, we embrace this shift. From simplifying agreement generation to streamlining how documents are created, edited, and signed, we’re employing AI throughout the document workflow to help organizations work smarter and faster. But AI also introduces its own set of problems.

AI has weaponized deception at scale. Deepfake video and audio are proliferating, and document and signature fraud have never been easier to execute. As a result, trust in digital systems is eroding across nearly every industry and rebuilding that trust requires moving beyond assumptions about identity to verifiable proof of who we are really interacting with.

The democratization and scale of identity fraud

The threats organizations face today look very different from those of just a few years ago. Fraudsters once needed complex operations to forge documents or impersonate executives, but AI has made deception cheap, accessible, and far more sophisticated.

Consumers lost a massive $27.2 billion to identity fraud in 2024. Businesses have fared just as badly, with the average organization reporting a loss of $7 million annually to identity fraud. And the problem is accelerating rapidly: financial institutions report a staggering 2,137% increase in deepfake fraud attempts over the last three years.

Recent real-world examples provide insight into how this type of fraud can take place and the impact it can have. In Italy earlier this year, business leaders were targeted via calls using an AI-cloned voice of the country's Defence Minister. The scammers claimed there were kidnapped Italian journalists and solicited urgent financial assistance, resulting in one high-profile businessman transferring nearly €1M.

In 2024 in Hong Kong, a finance worker attended what appeared to be a routine video conference with their company's CFO and senior staff — every other participant was an AI-generated deepfake. The realistic meeting convinced the worker to authorize 15 transactions totaling HK$200M to fraudulent accounts.

A foundational threat to business

The impact of AI-powered identity fraud goes beyond financial loss to an erosion of trust in the agreements organizations depend on. When they can no longer trust the critical point of agreement — the signing by each party involved — the effects are felt across the entire business.

Exposure to legal and compliance risk

For high-stakes transactions, a lack of (or insufficient) identity verification at the signing stage creates significant legal exposure:

• Organizations face increased legal disputes because they lack certainty about signature authenticity.
• Organizations in regulated industries face increasing scrutiny, and traditional eSignatures often result in gaps in compliance and security.
• The breakdown of identity assurance means digital business processes can no longer be relied on for the agreements that matter most.

Increased operational costs and delays

The lack of identity assurance leads directly to operational friction:

• Organizations face longer contract cycles due to manual checks, follow-up calls, or re-signing procedures required to address security concerns.
• There is a retreat to paper-based processes, which undermines the efficiency gains of digital transformation.
• Organizations experience higher operational costs due to extensive manual verification of signature authenticity for critical documents.

Sector-specific impacts

This pattern of vulnerability is consistent across industries:

• Financial services: Loan officers question whether loan applications are from real borrowers or synthetic identities.
• Professional services: Lawyers worry about the authenticity of client signatures on high-stakes agreements.
• Public sector: Administrators are unable to verify that citizens are who they claim to be when signing statutory declarations or applying for permits.

Identity must be proven, not assumed

AI makes identity fraud easier to create and harder to detect, which means trust can no longer rely on surface-level identity cues. The solution isn’t to pull back from digital processes, but to reinforce them with stronger assurance.

This is a critical challenge for agreements, where even a single mis-attributed identity can compromise compliance, security, or financial outcomes. Instead of relying on inbox access or device possession as proof of identity, organizations can adopt cryptographic methods that bind a verified, biometric-backed credential from the signer’s device directly to the specific document being approved.

This approach, exemplified by Verified Digital Signing, restores confidence in digital agreements without adding unnecessary friction. It anchors every approval to a verified individual and cryptographically links the signer, the document, and the moment in time, allowing trust, compliance, and efficiency to advance together.

Your guide to rebuilding digital trust

To understand how organizations are addressing the identity assurance gap and rebuilding trust in their digital workflows, download our white paper, Securing digital agreements in the age of AI deception.

Developed in partnership with MATTR, recognised pioneers in TrustTech, the white paper delivers real-world examples, industry insights, and a practical framework for strengthening digital trust in an AI-driven threat landscape.