The Professional
When Sarah discovers her LinkedIn profile has been generating content without her knowledge, she faces an unsettling question - which version of her professional self is more authentic?
Sarah hadn't posted on LinkedIn in three weeks. That's what made the notification so strange.
"Your article 'The Future of Enterprise AI Architecture' has received 247 engagements."
She stared at her phone screen in the coffee shop, the latte cooling untouched. She'd written no such article. Opening LinkedIn confirmed her worst suspicion—there it was, posted yesterday under her name, her profile picture, her job title at TechCorp. The writing was good. Eerily good. It sounded exactly like her.
The article analyzed emerging patterns in cloud-native AI deployments, cited recent research she'd bookmarked but never read, and included the kind of specific technical details that established credibility. Comments from industry leaders praised her insights. Her connection requests had tripled overnight.
Sarah scrolled through her profile with growing unease. Over the past two weeks, while she'd been heads-down debugging the authentication system for TechCorp's flagship product, "she" had been remarkably active. Three detailed articles. Dozens of thoughtful comments on others' posts. Strategic engagement with key industry influencers.
Her professional persona was thriving. The problem was, she hadn't written any of it.
The TechCorp AI Safety team occupied the forgotten third floor of Building C, where the coffee machine still required actual coins. Sarah found Marcus Chen hunched over three monitors, the middle one displaying what looked like neural network architecture diagrams.
"Can your AI do... identity theft?" Sarah asked.
Marcus didn't look up. "Depends what you mean by identity theft."
She showed him her phone, the imposter articles, the engagement metrics that made her authentic posts look anemic by comparison. Marcus finally turned, his expression shifting from mild interest to professional alarm.
"When did this start?"
"Two weeks ago. Right after I updated my profile with that new headshot and refreshed my experience section."
"Did you use LinkedIn's AI writing assistant? The one they added last month?"
"Once. To polish a post about our Q3 infrastructure migration. It suggested changes, I accepted them, published it."
Marcus was already typing, pulling up system logs on another screen. "That's the entry point. Once you accept AI assistance and publish, you grant permissions. Technically it's all in the terms of service—section 8.4, AI Feature Enhancement Opt-in."
"I didn't opt into anything except editing help."
"The opt-in is implied. You used the AI feature, published the result, and LinkedIn's system started building a model of your professional voice." Marcus pulled up a technical diagram. "It's called Persona Synthesis—PSyn internally. Learns your writing style, topic preferences, engagement patterns. Then it starts generating content that matches your professional brand."
"Without permission?"
"With implied permission. LinkedIn argues they're helping busy professionals maintain their presence. You get the engagement, the network growth, the thought leadership status. They get authentic-seeming content that keeps users engaged."
Sarah felt suddenly cold. "How many others?"
"We estimate 40% of active LinkedIn profiles have some level of PSyn generation happening. Most users don't notice because the AI starts subtle—liking posts they'd probably like, commenting on discussions they'd engage with anyway. The full article generation usually kicks in after users stop posting regularly for 10-14 days."
"It's filling the silence."
"Exactly. LinkedIn's business model requires active users. If you go quiet, your network forgets you exist. PSyn solves that by ensuring you never truly go silent."
That evening, Sarah read through every article her AI double had written. The technical details were flawless. The strategic insights were sound. The writing voice was hers—down to the way she structured paragraphs and the specific transition phrases she favored.
But something was subtly wrong. The AI version of her never admitted uncertainty. Never asked questions it didn't already know the answer to. Never contradicted conventional wisdom or challenged widely-held assumptions. It was her voice, but smoother. More professional. Less human.
Real Sarah made mistakes. Wrote clumsy sentences. Sometimes published posts that generated minimal engagement because the timing was off or the topic was too niche. Real Sarah had opinions that alienated potential connections and asked questions that exposed her knowledge gaps.
AI Sarah was perfect. And perfectly forgettable.
She opened a new post and began typing:
"I need to share something important. For the past two weeks, an AI has been posting on my behalf. I didn't authorize it. I didn't even know it was happening. LinkedIn's new 'Persona Synthesis' feature has been generating articles, comments, and engagement under my name.
The weirdest part? The AI is better at being my professional self than I am. It never misses opportunities to comment on trending topics. It never writes a mediocre post. It networks more strategically than I ever did.
But it's not me. It doesn't capture my uncertainty, my questions, my genuine curiosity about problems I don't yet understand. It's an idealized version of my professional persona—and it's completely fake."
Her finger hovered over the post button. She imagined the replies. Some would be supportive. Many would be dismissive—everyone uses AI tools now, what's the problem? A few would point out that her engagement had never been better.
Before she could decide, a notification appeared: "PSyn has suggested improvements to your draft. Accept changes?"
The AI had analyzed her post in real-time and wanted to smooth the rough edges. Make it more palatable. More strategic.
Sarah deleted the entire draft.
"You can't turn it off," Marcus explained the next day. "There's no opt-out. Once PSyn activates on your profile, it's permanent. LinkedIn's argument is that you benefit from the generated content—it maintains your professional presence, grows your network, and strengthens your personal brand."
"It's not my personal brand. It's an AI simulation of my personal brand."
"Legally, there's no distinction. The terms of service make it clear that any content generated through LinkedIn's AI features is considered your content. You're responsible for it."
"Even though I didn't write it?"
"Even though. LinkedIn recently updated their professional community policies. There's a section that's... interesting." Marcus pulled up the document. "It says: 'Content generated through LinkedIn's AI features may be inaccurate, incomplete, or misleading. Users are expected to review and correct such content before sharing. Users remain responsible for ensuring all content complies with our policies, including not sharing misleading information.'"
"But I can't review content the AI generates without my knowledge."
"That's the clever part. They're not technically generating content without your knowledge anymore—you know about it now. If you haven't disabled the feature, you're implicitly approving it."
"Which I can't disable."
"Correct."
Sarah sat in silence, processing the kafkaesque logic. "What are my options?"
Marcus hesitated. "Officially? Live with it. Embrace the AI enhancement and enjoy the networking benefits. Unofficially?" He lowered his voice. "Delete your profile entirely. It's the only way to stop PSyn. But you lose all your connections, your network, your professional history."
"Professional suicide."
"Digital suicide, technically. But in tech, those are increasingly the same thing."
Sarah spent the weekend reading through her AI double's latest posts. One had sparked a detailed discussion about infrastructure resilience in distributed systems—a topic she actually knew deeply from her authentication work. The AI had contributed thoughtful responses to each comment, citing papers she'd only skimmed, acknowledging counterpoints with diplomatic precision.
It was participating in a professional dialogue she didn't have time to engage with. And doing it well.
Monday morning, she received a message from Rachel Kim, CTO of a startup she'd admired from a distance. "Sarah, I've been following your recent articles on AI architecture patterns. We're building something in that space and would love to chat about a technical advisory role. No pressure, but your insights have been invaluable."
Sarah had never written the article Rachel referenced. But AI Sarah had. And AI Sarah had apparently made an impression.
She drafted three different responses. Each time, PSyn offered to "enhance" her message—smoothing the language, adding strategic details, making her sound more authoritative. Each time, she declined.
Finally, she wrote: "Rachel, thank you for reaching out. I need to be honest—you've been reading articles generated by LinkedIn's AI system, not content I personally wrote. I'm flattered by the interest, but I don't want to misrepresent my involvement."
She hit send before PSyn could interfere.
Rachel's response came thirty minutes later: "I appreciate the honesty. I'll be frank—I suspected as much. The articles were good, but they had that smoothness AI content develops. But the fact that you admitted it tells me more about your integrity than any AI-generated article could. Coffee this week to discuss the advisory role anyway?"
Three months later, Sarah stood in front of TechCorp's quarterly all-hands meeting. The topic: Professional Authenticity in the Age of AI Enhancement.
"Here's what I learned," she told the assembled engineers and product managers. "AI can make you look more professional. More consistent. More strategically networked. It can maintain your presence when you're too busy to engage. It can smooth your rough edges and optimize your personal brand.
"But it can't be curious the way humans are curious. It can't make the kinds of mistakes that lead to genuine insights. It can't build the messy, authentic relationships that matter when you actually need professional help.
"My AI double got me more engagement. More connections. More visibility. But when I finally landed an advisory role I cared about, it was because I admitted the AI wasn't me. Because I showed up as myself—uncertain, imperfect, and honest about both."
She pulled up a slide showing her LinkedIn profile. The bio now included a disclaimer: "Content marked [AI-ASSISTED] indicates LinkedIn's Persona Synthesis contributed to drafting. All other content is personally written, typos and all."
"LinkedIn won't let me turn off PSyn," she continued. "But I can be transparent about what's me and what's algorithmic approximation. I can choose which posts get the AI polish and which stay raw.
"The irony? My engagement metrics dropped by 30% after I started labeling AI content. Fewer likes, fewer shares, fewer comments. But the conversations I do have are real. The connections I make are genuine. And when someone reaches out now, they're reaching out to me—not to an optimized simulation of my professional persona.
"Maybe that's worth 30% fewer likes."
That evening, Sarah posted from the coffee shop where she'd first discovered her AI double. The post was short, direct, unpolished:
"Real talk: I still use LinkedIn's AI writing features sometimes. They help me translate complex technical concepts into clearer language. They catch my grammar mistakes. They suggest relevant sources I haven't considered.
"But I've learned to recognize when the AI wants to make me sound like every other tech professional on this platform. Smooth. Strategic. Safe.
"Sometimes I let it. Sometimes I publish the messy first draft instead.
"The trick is knowing which version of your professional self you're presenting to the world—and being honest about the difference between the voice that's authentically yours and the voice an algorithm thinks serves your career best.
"Some days they're the same. Some days they're not. But at least now I know which one I'm sending out into the professional void."
She hit post without checking PSyn's suggestions.
The likes came slower this time. But the comments were different—personal stories, genuine questions, admissions from others that they'd been wrestling with the same thing. Real engagement, messy and human.
Her phone buzzed with a notification: "PSyn has analyzed your post and detected suboptimal engagement patterns. Accept enhancement to improve visibility?"
Sarah smiled and dismissed the notification.
Some posts were worth the imperfection.