AiThority Interview with Valentin Vasilyev, Co-founder and CTO of Fingerprint

Valentin Vasilyev, Co-founder and CTO of Fingerprint chats about the Agentic AI ecosystem and how it’s transforming businesses at the core in this AiThority interview:

_____________

Hi Valentin, take us through a day at work, what does that look like as Fingerprint’s co-founder? What inspired the platform?

I’m a fan of remote and asynchronous work. Most of my day is spent reading and summarizing information from emails, Slack, sales calls transcripts etc. I usually have 2-3 short meetings per day with team members and sometimes with customers. My favorite tool is Notion, where I prefer to keep all my information and create short documents that help me put my thoughts in order.

I started the open source library, FingerprintJS, as a hobby while I was working as a JavaScript developer a couple of startups ago. The project became very popular quickly with thousands of websites using it. Over time, I realized that there was a real demand for a solution like Fingerprint that no other company was offering. So when Dan left the previous company, he asked me if he could work with me on the commercial version of Fingerprint, and that brings us to where we are now.

We’d love the top highlights on your latest AI agents launch: how does this new ecosystem benefit end users?

This new ecosystem gives organizations the ability to safely enable AI-driven workflows, from enterprise automation to e-commerce, without impacting the user experience. Companies can now identify every website visitor–whether it’s a malicious bot, an authorized agent, or a human—and can customize controls, rather than relying on a generic “block all bots” approach.  Authorized AI agents, such as those built on platforms like Manus and Browserbase, can access permissioned environments to analyze data without being flagged as a threat. Increased efficiency is a key benefit, as these agents automate tasks traditionally managed by human teams in order to streamline operations. For example, e-commerce and fintech companies can allow AI agents to facilitate transactions, creating a frictionless path for buyers and account holders while maintaining robust defenses against account takeover and payment fraud.

Also Read: AiThority Interview With Arun Subramaniyan, Founder & CEO, Articul8 AI

Why should more brands optimize how they identify AI traffic: what top best practices come to mind?

As AI adoption rapidly increases and agents account for a growing share of automated web traffic, enterprises need smarter and faster ways to parse through which traffic is legitimate versus potentially malicious. It has become more apparent that brands must shift away from legacy approaches that treat all automation as a threat, which breaks legitimate workflows while leaving businesses exposed and revenue on the table. The top best practices that come to mind are to replace legacy defenses with adaptive solutions that can detect AI-powered attacks, implement privacy-compliant persistent identification methods, and build layered defense strategies. These practices are especially timely in an era where it is becoming increasingly difficult to dodge bad actors. For example, as e-commerce sites open their platforms to verified AI agents, they are also creating more opportunities for malicious bots to exploit vulnerabilities. Companies that fail to adopt foundational infrastructure that intelligently filters legitimate traffic from threats will either face a sharp decline in conversions and rising false positives by blocking authorized traffic, or they will be forced to accept high-risk transactions.

What top myths surrounding the state and future of agentic AI would you like to bust in this conversation?

The myth that consumers are going to be the drivers of agentic AI adoption. Agentic commerce is a hot topic, but it largely depends on consumers abandoning their old ways of buying things online and switching to agents to do shopping for them. I don’t think this is going to happen fast, and we’ll need a few years to see consumer habits change.

Passkeys are a great example about consumer habits being slow to change. WebAuthn became standard in 2018. Passkeys are better than traditional passwords in every way. 99% of devices support passkey logins. Yet only 25% of consumers use them. This shows how much the power of habit influences consumers’ slow adoption of new technology.

In contrast, a much faster transformation is going to happen in the enterprise sector. Many types of tasks that were manually completed by employees will be automated with AI. Workflow automation across a heterogeneous set of tasks — such as customer support, CRM, sales operations, and coding — will likely be the first major enterprise domain to see broad AI tool adoption.

Five of the biggest AI innovations and trends you’re most excited about?

I have 3 I’m really excited about.

1. The number one thing I’m waiting for is when AI tooling becomes cheaper. Currently, AI usage is largely subsidized by massive investments, including from venture capitalists and government sovereign funds. When it gets to the point where revenue generation is a #1 priority, many users will be priced out, slowing down adoption and the rate of innovation. But AI shouldn’t be only for consumers in OECD countries. It should be available to all, such as schools in areas that may currently not have access to the internet — AI models could be made locally available to make education more accessible to everyone.
2. Specialized AI tools. We don’t always need massive AI chatbots that are best at everything at once. I hope to see faster, leaner, cheaper specialized AI tools that connect systems and act as key enablers in their domains. AI lawyers, AI doctors, AI judges (at sporting events) are all possible. To trust them and better integrate them into our everyday lives, they need to be better in their domains, inherently safer and more protected from abuse, hallucinate less, and cost less.
3. Safer AI. Safe in all meanings of the word. Safe from external attacks. Safe from social engineering. Safe in terms of ethical standards. Safe in dealing with personal information. Safe from political manipulation. Safe from inherent biases. Safe for humanity.

Also Read: Cheap and Fast: The Strategy of LLM Cascading (Frugal GPT)

[To share your insights with us, please write to psen@itechseries.com]