Agentic AI holds gigantic potential to improve and challenge ID verification

AI agents have entered the tech hype cycle, with significant implications for online payments and other digital interactions, and new requirements for digital ID systems. Businesses and individuals are expected to interact through AI agents integrated with digital wallets for identity verification.

Agentic AI is already changing identity verification, IDVerse argues in a blog post, and not just by generating demand to differentiate people from their AI representatives.

IDV Intelligence applies fraud detection proactively, and tailors verification with risk-based decision-making, IDVerse says. It also continuously refines itself based on the outcomes of verifications. Agentic AI allows ID verification systems to detect and block deepfakes before they show up in fraud attempts, adjust verification steps or reduce wait times for trusted users, and scale instantly to handle demand spikes.

IDV Intelligence is built to process data on users’ devices, and provide real-time fraud protection in compliance with global privacy laws. At the same time, the improvement over time that it makes results in fraud prevention, not just detection, while reducing false positives.

These advantages are largely a product of agentic AI’s ability to move beyond pre-set rules, IDVerse SVP of Enterprise Product Management Emily Hendley writes.

AI agent breakthrough ahead

The global market for agentic AI is expected to grow at a compound annual growth rate of over 43 percent, and near $200 billion by 2034, Peak VP of Strategy and Partnerships Chris Ashley says.

AI agents act independently, Ashley emphasizes. They don’t just respond to prompts, but take action, led by the large language model (LLM) “brain” of the agent. The use of tools from search engine to code-builders “is where agents really shine,” he says.

Ashley discussed what differentiates AI agents from each other in the market, and the ways they can help businesses, including several lofty claims of productivity and efficiency gains by enterprises. Realizing those gains is not simply a matter of flipping a switch to turn on AI agents though, Ashley notes. Testing, iterating, fine tuning and more are necessary to go from capability to reliability.

As businesses familiarize themselves with the potential benefits agentic AI can deliver, investment capital is flowing, and the AI developers are entering the market.

Market shaping up

San Francisco-based startup Human.org has raised $7.3 million to develop and market its blockchain infrastructure layer to support interactions between verified AI agents and real people, and its “Agent ID” authorization protocol.

Human.org’s infrastructure layer is offered in the form of a Human Network platform, and its portfolio also includes the native Human Dollar cryptocurrency and the Human App, a digital wallet which provides proof of humanity. It does so through a combination of face biometrics, “social graph, and ‘notarized credentials.’”

Twenty-three-year-old Founder Kirill Avery says use cases for the Human Network and Human App include self-sovereign identity (SSI) for interacting with organizations across a wide range of sectors.

Avery is aiming to keep bot accounts to less than 0.01 percent of the total Human Network. The company plans to use the funding to continue developing its technology. More details on the verification process are expected by mid-2024.

OwnID is adding native identity support for AI agents to its platform, for what the company calls “Customer and Agency Identity Management (CAIM).”

The digital identity solution is comprised of agentic access control system AgentLogin, a dynamic policy and consent enforcement layer, AgentGuard, and a system that creates a comprehensive audit trail, AgentTrace. The FIDO2-certified passkey-provider uses native device biometrics to authenticate humans on its platform.

OwnID CEO Dor Shany suggests AI agents “from personal shopping bots to enterprise assistants,” will soon represent millions of users.

Combatting AI agents is one of the stated purposes of Self Protocol, which provides digital identity infrastructure that uses zero-knowledge proofs (ZKP) to make token launches resistant to Sybill attacks.

The protocol has launched following the acquisition of OpenPassport by Self Labs. It can also be used for  privacy-preserving identity verification, location or age verification and DeFi integrations, according to a company blog post.

Article Topics

AI agents  |  digital identity  |  fraud prevention  |  Human.org  |  identity verification  |  IDVerse  |  OwnID  |  Self Labs  |  selfie biometrics