An AI medical tool is now used by two out of every three doctors in the United States

There is a company called OpenEvidence that most people have never heard of.

It does not make a chatbot you can play with. It does not generate images or write poems. It does one narrow, unglamorous thing:

Is Your Brand Showing Up in AI Search?

Most brands are invisible to AI, and they don't even know it.

The problem isn’t just about traffic. It’s about trust. When AI engines recommend your competitors over you, it compounds, as buyers form impressions about your brand before they ever reach your website.

Conductor's enterprise AEO platform solves that exact problem by giving marketing teams full visibility and control over how their brand appears across AI engines. To show you what that looks like for your brand, they're offering a free AI Search Visibility Report: a real-time snapshot of your AI presence, no setup or login required.

The report shows you:

• AI Market Share: How often your brand is mentioned and cited versus your competitors

• Funnel Visibility: Where you appear across prompts based on intent (Education, Recommendation, and more)

• Top Cited Content: Which pages AI engines already reference from your site

• Brand Sentiment: Whether AI is describing you positively, negatively, or not at all

AI search isn’t a fad. It’s how your customers find vendors, validate decisions, and shortlist solutions.

The challenge is ensuring your brand is part of those answers.

Get your free AI Visibility Report

It answers medical questions for doctors. Which drug interacts badly with which other drug. What the latest research says about a specific treatment. What dosage is recommended for a patient with a particular condition. Simple questions, asked a hundred times a day, that used to require a physician to stop, search, and verify.

As of this week, roughly 650,000 doctors in the United States use it — about two thirds of all practising physicians in the country. Another 1.2 million doctors use it internationally. That is not a product in beta testing. That is infrastructure.

Why this matters for you

OpenEvidence is not famous. It does not trend on social media. But it is probably worth more than most AI tools you have heard of — because it owns a daily habit inside one of the most trusted professions on earth. The lesson is not about medicine. It is about what happens when AI solves a real, recurring problem for a specific group of people who have no time to waste.

Boring industries, extraordinary leverage

This is a pattern worth paying attention to. The AI tools grabbing headlines — the image generators, the writing assistants, the chatbots — are competing in the most crowded markets imaginable. Everyone can use them. That also means everyone offers them. Margins are thin, switching is easy, and loyalty is low.

The quieter story is happening in industries that feel dull from the outside but are enormous on the inside. Healthcare. Logistics. Property management. Accounting. Legal research. Agriculture. These sectors share three things: they are full of repetitive, time-consuming tasks; the people doing those tasks are highly trained and deeply underserved by existing software; and they have serious money to spend when something actually works.

A tool that saves a doctor ten minutes per patient, across a hundred patients a week, is not a convenience. It is, in practical terms, the equivalent of hiring an extra staff member. That kind of value is easy to justify paying for — and hard to walk away from once it is embedded in the daily routine.

"The AI tools that will last are not the ones that impress people at dinner parties. They are the ones that professionals cannot imagine their morning without."

What this means if you are building something

The OpenEvidence story carries a direct lesson for anyone thinking about a side business or an AI product. The instinct is usually to go broad — build something for everyone, and let the market decide. That instinct is almost always wrong.

Going broad means competing with OpenAI, Google, and every well-funded startup on the planet. Going narrow means competing with almost nobody, serving people who have a specific pain, and building the kind of daily habit that makes your product feel less like software and more like a colleague.

The niche does not have to be medicine. It can be the local tax rules that confuse small business owners in Nigeria. The visa paperwork that overwhelms skilled workers moving from the Philippines to the Gulf. The inventory management headaches that keep small retailers in Jakarta up at night. Every one of those is a real problem, felt by real people, in markets that most Western AI companies have never thought about.

How to find your version of this

01Start with a profession, not a product. Pick a job where people do the same tasks repeatedly — nurses, property agents, logistics coordinators, accountants. Ask: what do they do every day that is slow, annoying, and solvable with information?

02Build the habit before the business. A weekly newsletter, a prompt library, a free template — something that becomes part of their Tuesday morning. Once they rely on you, the paid product is a natural next step, not a cold pitch.

03Think local. The most underserved markets right now are not in San Francisco. They are in cities where English is a second language, regulations are complex, and almost no AI product has been built with local context in mind. That gap is an opportunity.

04Ignore the feature race. OpenEvidence did not win by having more features than its competitors. It won by being reliable, specific, and trusted. Trust is the moat. Build for that first.

Two thirds of American doctors now use an AI tool that most people cannot name. That anonymity is not a weakness. It is the whole point. The best AI businesses of the next decade may never trend on Twitter. They will just quietly become impossible to work without. ■

Top News This Week

Americas

Sam Altman and Elon Musk face each other in court over who really owns OpenAI's future

Reuters / The Guardian  ·  May 13, 2026  ·  Altman testified that Musk once pushed to hand control of OpenAI to his children and demanded total ownership. The trial, tied to Musk's claim that OpenAI betrayed its non-profit mission, now overlaps directly with OpenAI's push toward a reported $1 trillion public listing.

76% of large companies have now created a "Chief AI Officer" role — up from 26% last year

IBM / CNBC  ·  May 12, 2026  ·  A new IBM survey of 2,000+ organisations finds AI has arrived in the boardroom. The harder question: most companies still cite culture, not technology, as the main barrier to adoption.

OpenAI backs a global AI governance body modelled on the nuclear watchdog — and wants China inside it

Bloomberg  ·  May 13, 2026  ·  OpenAI told US lawmakers it supports an International Atomic Energy Agency-style body for AI, with the US in the lead and China as a member. A significant shift from the America-first framing that has dominated AI policy.

APAC

Moonshot AI raises $2 billion, valuing China's answer to OpenAI at $20 billion

Bloomberg / TechStartups  ·  May 7, 2026  ·  The maker of the "Kimi" AI models has now raised nearly $4 billion in six months. Annual revenue recently passed $200 million, and its latest model is among the most widely used in China. The gap between US and Chinese AI labs is closing faster than most Western analysts expected.

Alibaba's AI and cloud business is finally making serious money — not just spending it

HIPTHER / May 13, 2026  ·  After years of heavy investment, Alibaba's AI division is beginning to show real revenue at scale. A signal that the infrastructure buildout phase in Asia is starting to pay off.

Europe & Africa

AI "vibe-coding" tools are leaking sensitive data — 380,000 files exposed by accident

MarketingProfs / May 8, 2026  ·  Security researchers found that tools like Lovable and Replit allow non-technical staff to build production software without understanding basic security. Healthcare records, financial data, and internal systems are among the exposed files. A reminder that AI making things easy also makes mistakes easier.

Google to launch a new Gemini model at I/O next week, reportedly near GPT-5.5 in capability

The Verge / Sources  ·  May 13, 2026  ·  The model wars continue. Google's upcoming release is expected to close the gap with OpenAI's latest, just as the two companies battle for developer loyalty and enterprise contracts.

This Week's Top AI Funding Rounds

Sources: Crunchbase, Reuters, Bloomberg, TechStartups, NBC News, The Guardian, MarketingProfs. Funding figures reflect publicly reported data as of May 14, 2026. The AI Edge is an independent newsletter. We do not accept paid placements.