AI Is Making The Internet’s Bot Problem Worse. This $2 Billion Startup Is On The Front Lines

Key Concepts

• AI-driven bots and their increasing prevalence on the internet.
• Challenges posed by AI to traditional bot detection methods.
• Identity verification as a crucial aspect of combating malicious AI activity.
• Persona's approach to identity verification using a blend of methods and machine learning.
• Customized verification processes ("flows") tailored to specific use cases and risk levels.

The Growing Bot Problem and the Rise of AI

The internet faces a significant challenge with the proliferation of bots, which currently account for approximately 50% of all internet traffic. This number is projected to surge to 90% by 2030 due to the increasing use of AI agents for automated online tasks. This poses problems for online retailers, banks, and social networks, who need to prevent malicious bot activity while allowing legitimate bots to function.

AI's Impact on Bot Detection

Traditional bot detection techniques, such as CAPTCHAs, are becoming ineffective against sophisticated AI models. AI also facilitates the creation of realistic voice and face spoofs, as well as fake IDs, making it harder to distinguish between humans and bots. Rick Song, co-founder and CEO of Persona, suggests that the focus should shift from simply identifying bots to understanding the intent behind the AI and the identity of the user controlling it.

Persona: An Identity Verification Platform

Persona, a San Francisco-based startup, offers ID authentication software to 3,000 companies, including OpenAI, LinkedIn, Etsy, Reddit, DoorDash, and Robinhood. The company employs a multi-faceted approach to verify user identities:

• ID Upload: Users are asked to upload a photo of their government-issued ID.
• Selfie/Video Verification: Users take a selfie or record a video while performing actions like turning their head or blinking.
• NFC Chip Scanning: Scanning the NFC chip embedded in passports.
• Liveness Tests: High-risk users may be required to hold up their ID and move their face to prove they are alive and the genuine owners of the ID.

Persona's machine learning models analyze various risk signals, including network information, location proximity to the ID, device interaction patterns, and online presence. According to Song, bot activity often exhibits mechanical characteristics, such as excessive copy-pasting, hesitations between inputs, and rhythmic behavior.

Funding and Valuation

In 2023, Persona raised $200 million in Series D funding led by Ribbit Capital and Founders Fund, with participation from existing investors. The company is now valued at $2 billion and has secured a total of $417 million in funding. In the previous year, Persona signed $100 million worth of annual contracts.

Customized Verification Flows

Persona's success is attributed to its ability to build customized verification processes, known as "flows," that are tailored to the specific needs of its customers and adjusted based on the risk level presented by the user. For example, the verification process for age verification for alcohol delivery differs from that for loan applications. Song emphasizes that "There's no one-size-fits-all for identity."

The Financial Impact of AI-Based Attacks

AI-based attacks result in significant financial losses for companies. US companies lose between $18 billion and $31 billion annually due to these attacks, according to a report by Imperva. Global losses from bot attacks, including server cost inflation, range from $68 billion to $116 billion. Bots can exploit incentives like referral credits, discounts, and promotions by creating fake accounts.

Conclusion

The rise of AI-driven bots presents a complex challenge to internet security and identity verification. Traditional bot detection methods are becoming obsolete, necessitating a shift towards understanding user intent and verifying identities. Persona's approach, which combines multiple verification methods, machine learning, and customized verification flows, offers a promising solution to combatting malicious AI activity and mitigating financial losses.