Austin Founder Launches Clare, an AI Desktop Assistant for People Overwhelmed by the Internet

Why it matters

Consumer security is still failing nontechnical users at the moment of decision. Clare's opportunity depends less on detection claims than on whether Ryan Merket can make protection simple enough for the people most targeted by scams.

For most technology companies, the pitch begins with a market opportunity.

For Ryan Merket, it began with his parents.

The Austin-based founder says he built Clare after watching older family members struggle with an internet that increasingly feels hostile to non-technical users. Phishing attempts arrive disguised as account alerts. Scam messages impersonate trusted brands. Routine online tasks often require navigating unfamiliar interfaces, security prompts, forms and notifications.

"Watching my parents navigate the internet honestly terrifies me," Merket wrote in a recent post on X while introducing the product.

The result is Clare, a desktop-based AI assistant currently operating in a private alpha. Rather than positioning itself as a cybersecurity tool, the company describes Clare as a patient digital helper designed for people who find modern computing confusing or overwhelming.

According to company materials, Clare appears on a user's desktop as a small purple orb that users can speak to in plain language. The assistant is intended to help with common digital tasks including email, Facebook, web browsing, online forms, banking websites, video calls and general computer navigation.

The company says Clare supports 38 languages and is designed to provide assistance without the judgment or frustration many users experience when asking friends or family members for technical help.

A Different Approach to Consumer Security

The broader consumer security market is hardly underserved.

Major technology platforms have spent years deploying spam filters, phishing detection systems, account recovery tools and security warnings. Browser vendors regularly block malicious websites. Email providers continuously improve scam detection.

Yet fraud remains pervasive.

What Clare appears to be targeting is not necessarily the technical problem of identifying malicious activity, but the human problem of uncertainty.

Many users still receive messages they cannot confidently evaluate. They hesitate before clicking links, opening attachments or entering credentials. They often know that scams exist but lack the confidence to determine what is legitimate in real time.

That gap between available security technology and actual user confidence may represent Clare's opening.

Rather than adding another warning label or security dashboard, the product aims to create a conversational layer between users and the increasingly complex systems they interact with every day.

Trust Will Be the Critical Question

The company's public materials place significant emphasis on privacy and trust.

Clare is described as operating on the user's computer, and the company says it does not sell user data. Those assurances are likely to be central to adoption, particularly among the demographic groups the product appears designed to serve.

The challenge for any AI assistant operating near sensitive activities such as email, banking and account management is that trust must be earned before users are willing to delegate meaningful tasks.

That challenge is especially pronounced in a market where consumers are simultaneously being warned about AI-generated scams, impersonation attacks and increasingly sophisticated social engineering campaigns.

The product's success may ultimately depend less on the sophistication of its underlying models and more on whether users view it as a reliable guide during moments of uncertainty.

Why It Matters

The AI industry has spent the last several years racing toward more capable models. Much of the conversation has focused on coding assistants, enterprise automation and agentic workflows.

Clare represents a different thesis.

Instead of helping power users move faster, it attempts to help less technical users feel safer and more capable online.

That market is enormous, but historically difficult to serve. Consumers rarely buy security products until after something goes wrong, and many assistive technologies struggle to gain sustained engagement.

If Clare can successfully combine AI assistance, accessibility and trust, it could address a problem that remains largely unsolved despite decades of progress in consumer software: making the internet feel usable for everyone.

The company is currently accepting applications for its private alpha program.

Disclosure: Ryan Merket has a relationship with RuntimeWire.