When I was a young reporter covering the Afghan war, I spent Saturday afternoons watching buzkashi matches played by Afghan refugees in Peshawar, Pakistan.
Buzkashi, Persian for “goat-grabbing,” is Afghanistan’s national sport. Riders, known as chapandaz, compete to seize a goat carcass from the haram circle, carry it up and down the field, and drop it in the halal circle to score.
At the matches I watched in those refugee camps back in 1989, the formal structure was collective. The incentives were personal.
A successful run counted for the rider’s side. But the chapandaz who scored also received a cash prize. So when a rider seized the goat, everyone on the opposing team tried to take it from him.
So did his teammates.
In that particular version of the game, the side gained the point. The rider captured the reward.
It was easy to see why Western reporters seized on buzkashi as a go-to metaphor of Afghan politics. Alliances were real but provisional. Formal loyalties coexisted with personal ambition. Everyone might be advancing the same side, but possession changed the incentives.
This pattern is not unique to Afghanistan. It appears whenever an organization depends on individuals whose ambitions, audiences, and authority cannot be fully contained by the institution. That tension has always existed. AI is making it harder to ignore.
Elon Musk is an extreme case.
His persona is larger than any one of his companies. From Tesla to DOGE and SpaceX, his interventions feed a continuous spectacle that transfers attention from one venture to another. It can also impose reputational costs that no corporate communications function can contain.
Musk's business ventures don’t simply confer authority on him. They participate in an authority system organized around him.
Musk eclipses the institution. Altman personifies it. Amodei articulates its worldview.
Sam Altman represents a different model. He has become the human proxy through which much of the world evaluates OpenAI’s ambitions, judgment, and trustworthiness. Debates about whether the institution should be trusted frequently collapse into debates about whether Altman should be trusted.
Dario Amodei offers yet another variation. His long-form arguments do more than promote Anthropic. They propose ways of understanding what advanced AI could mean, which risks deserve attention, and what choices society faces.
When a company claims authority over a transformative future, audiences look for a person to hold responsible for the claim. The company builds the technology. The leader explains the stakes.
Musk eclipses the institution. Altman personifies it. Amodei articulates its worldview.
AI didn't create celebrity CEOs. It changed the economics of authority. To understand why, it helps to look at how authority now travels.
A chief scientist’s posts attract more serious attention than the corporate newsroom. A founder’s newsletter defines the category more clearly than the company website. An engineer becomes the market’s preferred interpreter of the product their employer sells.
The authority may benefit the company, but is attached to a person. The rider can change teams, strike out independently, or carry the audience somewhere the institution never intended to go.
Digital platforms have been shifting influence toward individuals for years. The pattern has only deepened since the advent of generative AI.
A 2023 analysis by the B2B marketing firm Refine Labs compared the LinkedIn performance of seven employee profiles with the company’s own page. The personal profiles generated 2.75 times as many impressions and five times as much engagement.
It was a single-company study, not a universal law. But it quantified a familiar pattern: company pages certify the institution, while people conduct much of the actual conversation.
Engagement is not the same as trust. People engage with outrage and spectacle as readily as expertise. But engagement shows where audiences are directing attention and granting opportunities for influence.
Repeated exposure creates familiarity. Familiarity can become reliance. Over time, the individual voice becomes the lens through which the institution is understood.
The pattern extends beyond business media. In a 2024 study, Pew Research Center found that 21 percent of American adults regularly got news from social-media influencers. Among adults under 30, the figure was 37 percent. Of those who regularly consumed news from influencers, 65 percent said those individuals helped them understand current events better.
People aren’t necessarily abandoning institutions. Many influencers draw on reporting, research, or credentials produced by established organizations. But audiences increasingly prefer to source that knowledge through an identifiable person who selects, interprets, and explains it for them.
Organizations own the assets. Often, individuals carry the authority.
Once upon a time, organizations controlled the machinery of authoritative expression.
They deployed researchers, editors, designers, PR teams, agencies. They controlled publishing channels. They could produce polished arguments, persuasive reports, professional graphics, and polished executive voices.
Individuals outside those systems could speak, but few could match their reach or production quality.
Engagement is not the same as trust. People engage with outrage and spectacle as readily as expertise.
Generative AI has weakened that advantage. A company, a chief executive, an employee, a consultant, a customer, an independent expert, and an anonymous account can now produce fluent arguments, polished frameworks, credible-looking evidence, and professional media at negligible cost.
AI hasn’t abolished institutions. It has palpably narrowed their advantage in authoritative expression.
That doesn’t mean every voice carries equal weight. Companies still have the advantage when it comes to capital, customers, proprietary data, operating capability, distribution, and accountability. But the surface distinctions have narrowed. The institutional voice no longer arrives automatically dressed in greater authority than the individual one.
AI accelerates this shift in two ways.
First, it gives more people the means to produce polished public expression. Second, by making polished expression abundant, it increases the value of signs that a real person stands behind the words.
Sprout Social’s 2026 research found that human-generated content was the top priority consumers wanted brands to pursue on social media. In a related survey, 55 percent of social users said they were more likely to trust brands that published human-generated content, rising to roughly two-thirds among Gen Z and millennials.
The more language can be generated without a speaker, the more audiences want to know who is speaking.
The finding is easy to misread as a rejection of AI. It’s more interesting than that. Audiences don’t necessarily object to AI being used in research, production, editing, or distribution. What they resist is content that appears to have no accountable intelligence behind it.
The more language can be generated without a speaker, the more audiences want to know who is speaking.
What has this person seen? What do they understand from experience? What evidence can they produce? What judgment are they prepared to defend?
These questions pull authority toward individuals because individuals can answer them in ways that logos cannot.
Individual bylines are necessary, but insufficient. AI can manufacture executive posts, founder letters, opinion pieces, and intimate-sounding reflections almost as easily as corporate copy.
A first-person pronoun hardly proves first-person judgment.
This is where institutions re-enter the argument.
The individual may perform in the market, but institutions provide the evidence that translates performance into authority: original research, customer outcomes, technical achievement, operating experience, and a record of consequential decisions.
With that foundation, the individual becomes a credible interpreter of the business.
For organizations, this creates an apparent contradiction.
Audiences increasingly prefer to hear from people. Yet the authority of those people often depends on assets the organization has built: data, products, customers, laboratories, engineering teams, or privileged visibility into a changing market.
The individual voice and the institutional base are not alternatives. They perform different functions.
The individual attracts attention, exercises judgment, and makes the institution intelligible. The institution supplies proof, continuity, resources, and accountability.
B2B buying research shows why that combination matters. In 2024, Edelman and LinkedIn found that 73 percent of decision-makers considered thought leadership a more trustworthy way to assess a company’s capabilities than marketing materials or product sheets. Their 2025 study found that 53 percent believed strong thought leadership made brand recognition matter less.
This research is usually presented as an argument for better content. But it points to a deeper shift in authority.
Buyers aren’t satisfied with an organization declaring what it can do. They want access to the thinking of the people inside it. They want to see how those people interpret the market, distinguish signal from noise, and connect the company’s capabilities to problems that matter.
The most valuable individual authority is not merely adjacent to the business. It reveals something the business uniquely knows, sees, or can prove.
At the same time, thought leadership works as evidence of corporate capability only when the thinking is visibly connected to the organization. An executive who accumulates a large following by commenting on every fashionable subject may build personal reach while leaving the company’s position no clearer.
The conventional corporate response to prominent individual voices is control.
Define the approved narrative. Train the spokespeople. Route public statements through comms. Reduce variation. Keep the rider galloping alongside the team.
That approach made more sense when organizations controlled the channels through which authority was produced and distributed. It works less well when every executive and employee has access to publishing tools, generative systems, and a potentially global audience.
Tight control often produces language that is accurate, polished, and easy to ignore. At the other extreme, a collection of strong personal brands may build visibility without adding up to a coherent market position.
The alternative is not to make everyone say the same thing. It is to create a strong enough shared frame that different people can say different things while strengthening the same market position.
That requires more than a messaging document. It requires a body of evidence people can draw upon, a clear account of the change the company sees, and a common understanding of the category it seeks to shape.
The chief executive, researcher, engineer, and customer should not sound interchangeable. They should sound as though they understand the same market from different positions within it.
Their individual authority then becomes a distributed system rather than a collection of freelance performances.
At those buzkashi matches in Peshawar, the energy came from the fact that every chapandaz wanted possession. Personal ambition was not a defect in the game. It was what drove the riders down the field.
Organizations need some of that ambition too. They need leaders and experts who want to make arguments, develop audiences, and become authoritative in their own right.
The strategic task is to connect that personal reward to a larger score.
AI has made it easier for almost anyone to sound like an institution. At the same time, digital audiences increasingly expect institutions to speak through recognizable people.
Individuals possess more expressive power. Organizations retain the assets that can turn expression into proof. Platforms determine which voices receive attention, while audiences decide which voices become trusted interpreters.
Elon Musk shows what happens when the rider eclipses the team. Sam Altman shows how completely institutional trust can become attached to one human proxy. Dario Amodei shows how an executive’s ideas can make a company’s worldview legible.
Most organizations will never operate at that scale. But they face the same underlying problem: institutions own the assets, while individuals increasingly carry the authority.
The challenge is making every rider’s authority count for the team.