Brian Long Built Two Marketing Companies. Then a Fake Brian Texted His Team.

A new employee at Attentive received a text from the boss. The request sounded plausible enough, and the employee made a significant business purchase. There was only one flaw: the boss was not Brian Long. When the incident reached the employee's manager, the impersonator tried again. The manager was fooled too. One fake founder had managed a two-person winning streak.

Long could have filed the episode under workplace embarrassment, a drawer already crowded at most companies. Instead, it became an operating clue. He had spent seven years as Attentive's chief executive, helping turn a company begun in his living room into a messaging platform with more than 8,000 customers. Its business was communication. Yet two people inside it had been beaten by a common text message wearing a familiar name.

The awkward symmetry was hard to miss. Long had built software that helped brands reach people on their phones. Now the phone had become the scene of the crime.

“I couldn't believe there wasn't a better way to stop these attacks.”Brian Long, recalling the Attentive impersonation incident

A career spent following attention

Adaptive Security looks like a sharp turn from Long's earlier work. Seen from a few paces back, the road is straighter. His companies have repeatedly occupied the narrow strip where software meets human attention.

Before founding companies, Long worked in sales roles at CNET and led mobile products at Pontiflex, later called Crossboard Mobile. In 2012 he co-founded TapCommerce with Andrew Jones. The company helped app marketers bring users back after installation - a useful problem in the first great rush toward mobile apps. Twitter acquired it in 2014 for a reported sum above $100 million. Long stayed as a senior director, gaining the peculiar education that follows when a startup is swallowed by a platform.

Two years later, Long and Jones started Attentive. The wager was that shoppers would not install an app for every retailer, but might accept a useful text relationship. Its deceptively modest two-tap sign-up compressed consent into a simple mobile action. By the time Long moved from CEO to executive chairman in 2023, the company had become a substantial enterprise, with 1,000-plus U.S. full-time employees and $500 million in annual revenue, according to the account he later published.

The third act began with observation rather than inheritance. Long did not present himself as a lifelong security specialist. In 2023, he and the team spoke with hundreds of chief information security officers. One concern kept returning: AI-powered social engineering. The old misspelled phishing note was gaining grammar, context, patience and, increasingly, a voice.

There was another education hiding in Attentive's boardroom. Long has recalled reading the security team's quarterly updates and seeing multiple material incidents at a company of roughly a thousand people. Attentive was large, but hardly a household name. If this much trouble could accumulate inside one ordinary technology company, he wondered, how much was going unreported everywhere else?

Cybersecurity, he later observed, is opaque by custom. Companies tend to speak only when an incident is too disruptive to hide. Long borrowed a pest-control analogy: seeing one mouse suggests many more are inside the wall. Adaptive's early customer conversations gave him a wider keyhole. People would describe what had happened last week, even when none of it appeared in public. The market research was not a stack of cheerful survey responses. It was a confidential chorus of near misses.

The boss has entered the uncanny valley

A modern impersonation can be assembled from ordinary corporate debris. A headshot supplies a face. A conference clip supplies a voice. LinkedIn supplies colleagues, reporting lines and recent promotions. An attacker can add an urgent request, then carry the conversation from email to text to a live call. Each detail makes the lie feel less like a lie.

Long describes the change as a move beyond the deepfake toward the “deepfake persona.” The counterfeit is no longer merely an image or sound. It has a small biography. It knows who reports to whom. It can improvise. In a podcast interview, he warned that a capable attacker could use agents to run many such efforts at once, giving fraud some of the same scale economics that founders celebrate in legitimate software.

Adaptive's response is rehearsal. The platform can run simulated attacks through email, SMS and voice, using the kinds of public information that a real attacker might collect. When an employee fails a simulation, that event can alter a risk score and trigger training related to the mistake. The aim is to replace the annual compliance fog with a specific memory: the call sounded like my chief executive, but the request violated our process.

The company has widened the brief since its public launch. Its 2026 releases included email security, an AI governance product for seeing unsanctioned tools, a content studio, branded training and AutoPhish. The expansion reflects Long's stated habit of listening closely to customers and updating the product. It also reveals how quickly a tidy category called “security awareness” spills into the rest of the organization.

One new product also circles back to the unglamorous problem of employee attention. Branded Training lets a company apply its own visual identity inside lessons, not merely paste a logo around the edges. Long's explanation is behavioral: generic colors and stock imagery tell employees that a module came from somewhere else. It feels outsourced before the first question appears. In security, even typefaces can become part of the trust problem.

The repeat founder's unfairly reusable trick

Long has been candid about entering unfamiliar territory. Asked whether founders should build only where they already possess expertise, he rejected the restriction. Large markets can be learned, he argued, if a founder is willing to immerse himself, meet people and recruit well. Familiarity helps with credibility and fundraising, but it can also harden into stale assumptions. “Knowledge ages really quickly,” he said.

That line is more revealing than the standard mythology of the repeat founder, who supposedly walks into the next company carrying a leather case of answers. Long's portable advantage appears closer to a method: notice a changing channel, interview the people living with the change, and make the troublesome action easier to understand.

There is a useful tension here. Adaptive uses AI to generate the kinds of deceptions that AI has made cheap. It studies exposure in order to reduce exposure. A convincing simulated call must be unsettling enough to teach, yet controlled enough to preserve trust. Long has acknowledged that security tests can annoy people, sometimes spectacularly. The product line is therefore both technical and theatrical: create a safe performance of betrayal without making the cast hate the director.

“Cybersecurity now begins with people, not just infrastructure.”Brian Long, after the OpenAI Startup Fund's follow-on investment

A small pause, professionally installed

Adaptive's $81 million Series B, announced in December 2025, brought in Bain Capital Ventures, NVIDIA, Capital One Ventures and Citi Ventures alongside existing investors. The company said it was already protecting more than 500 enterprise customers. Capital can buy engineers and distribution. It cannot slow the models on the other side, which is why Long speaks about the problem with a blend of urgency and routine maintenance. New capabilities appear every few weeks. The defenders must keep playing with them.

The funding story has its own repeat-founder rhyme. Bain had backed Long at TapCommerce and later led Attentive's first major round. When Attentive emerged from stealth in 2018, Long joked that investor Scott Friend had been a strong board member the first time, so why not repeat the experience? Years later, Bain led Adaptive's Series B. Venture capital prefers novelty in public and pattern recognition in private.

His most memorable public advice is wonderfully unceremonious: “Delete your voicemail.” A clean outgoing message can offer useful material for voice cloning. The tip will not solve synthetic fraud, but it captures Long's practical register. Grand threats are experienced through small surfaces - an inbox, a phone, a familiar face in a meeting window - and defenses must eventually arrive there too.

The aspiration behind Adaptive is larger. Long says the mission is to stop AI-powered social engineering. The day-to-day product is more modest and perhaps more believable: install a pause between recognition and obedience. Make a person wonder whether the urgent message deserves a second channel, a passcode or a call back. Turn suspicion from a personality defect into an organizational skill.

Long's career began by helping companies earn attention. It moved to helping brands enter the most intimate commercial channel, the text thread. Now he is working on the cost of that proximity. The closer software gets to us, the more valuable a counterfeit becomes. His third company is betting that trust can survive the counterfeits, provided people get to meet a few of them in rehearsal first.