Feature
In boardrooms across the globe, executives are making a bet that AI-powered personalization will unlock unprecedented employee productivity. The promise is seductive: Intelligent systems surface exactly what each employee needs, when they need it, cutting through information overload and boosting efficiency.
But as organizations rush to deploy these digital gatekeepers, a worrying pattern has emerged. The very systems designed to enhance digital collaboration and innovation at work may be dismantling them without anyone noticing.
The New, Invisible Workplace Silo: AI Personalization
AI-driven personalization offers mixed blessings, said Loren Absher, director, Americas lead — applied AI advisory at ISG. "AI acts like a digital gatekeeper, surfacing what it thinks you need and quietly hiding the rest," he explained.
These systems analyze employee roles, communication patterns, click history and team behavior to create highly curated information feeds. Yet Absher warns of an unintended consequence. "While the goal is relevance, the unintended outcome is that tangential or unstructured content — often the spark for breakthrough thinking — gets buried," he said.
This filtering effect creates what Absher calls a fundamental shift in organizational structure. "Over-personalization is the new silo — and it's algorithmically enforced. Traditional silos were built by org charts; today's silos are built by code." The implications extend far beyond missed emails or overlooked reports. As AI systems become more sophisticated at predicting what employees want to see, they simultaneously become more effective at hiding everything else.
Rackspace Senior Director Abhishek Agrawal has observed this phenomenon across multiple enterprise deployments. "Over-personalization can reduce cross-pollination of ideas," he noted, describing scenarios where "engineers may only see content from their immediate team, missing exposure to marketing insights that could spark product innovation."
What makes this particularly insidious is that the efficiency gains are immediately visible — employees get through their information feeds faster — while the innovation losses become apparent only over time.
The AI Efficiency Paradox
The cascading effects of these algorithmic barriers reveal themselves in unexpected ways. Laurel Long, principal behavioral scientist at CoachHub, shared an example many organizations will recognize: "A product team asks an AI tool to generate a roadmap tailored only to their team's past preferences and data,” she said. “The tool suggests improvements that mirror what the team has already done, but because it didn't pull in insights from customer support, sales or marketing, the roadmap misses breakthrough ideas."
The example highlights a crucial distinction between beneficial and harmful personalization, Long said. "If the way that content is conveyed is personalized, then I don't foresee a hindrance to cross-team collaboration or innovation. However, if what content is communicated is personalized, then it could hinder innovation by over-reliance on AI and not enough reliance on experts from other teams."
The difference is subtle but profound: personalizing the delivery of information maintains its breadth, while personalizing the selection of information narrows it.
The innovation implications become clearer when viewed through Absher's lens of organizational dynamics. "Innovation often comes from chance collisions — like overhearing a problem outside your domain. Over-personalization eliminates those collisions before they happen," he said. This connects directly to Long's roadmap example — the product team never had the chance for that "collision" with customer support insights because the AI system deemed them irrelevant.
But the damage extends beyond missed innovation opportunities. A more fundamental erosion is taking place: "When AI continually narrows what we see, employees start operating in algorithmic silos. This can lead to misaligned execution, duplicated work and fractured understanding across the organization," said Absher.
When the tools designed to reduce redundancy instead increase it by preventing teams from seeing each other's work, the efficiency paradox becomes clear.
The Effect of AI Bubbles on Culture
The cultural implications of this trend worry experts most. "Culture depends on shared context. AI bubbles quietly erode that common ground," Absher warned.
"A strong culture relies on shared narratives, language and values. When employees each see a different filtered slice of the company, that common foundation frays." Long builds on this concern, explaining how "personalized content treated as the sole source of truth" leads to employees becoming "increasingly locked into narrow perspectives, making it harder for knowledge to flow across teams."
If we take this analysis to its logical conclusion, the future will be one where: "leadership abdicates too much control, allowing AI to drive strategy solely in the name of optimization," said Digitate Chief Customer Officer Ugo Orsi.
In such scenarios, Orsi warned, "The organization runs the risk of empathy being lost, and the company may become entirely focused on efficiency." This connects back to Absher's opening observation about AI as gatekeeper: When the gatekeepers become too powerful, they reshape not just information flow but organizational culture.
A Balanced Approach to Personalization
No one is advocating for abandoning personalization entirely. Instead, they call for more thoughtful implementation.
"Personalization should be treated like seasoning: too little and work feels overwhelming; too much and it becomes blandly predictable,” Absher said. Finding the right balance requires what he calls "engineered serendipity" — systems that "inject calculated randomness, surfacing unexpected content that leads to fresh ideas."
For her part, Long advocates for a "multi-stream approach" where organizations, "treat personalization like one stream of information, not the only stream of information." She recommends systems that "highlight core resources tailored to an employee's role but also periodically share content from other teams, industries or disciplines." Similarly, Agrawal has observed organizations experimenting with "controlled randomness" in recommendation systems, which deliberately surface content outside employees' immediate scope to encourage discovery and lateral thinking.
The monitoring challenge these solutions create requires new approaches. "Treat AI filters like any business system — tracked, stress-tested and designed with opt-outs," Absher recommended. "AI is not magic; it is infrastructure."
"If teams begin working in silos, recycling the same ideas, or producing projects that lack cross-functional input, it's a sign that personalization may be narrowing perspectives," said Long, suggesting that businesses focus on outcomes rather than algorithms.
Questions to Check Your Personalization Settings
The key questions leaders should ask relate directly to the cultural concerns raised earlier, said Long:
- Are workflows becoming more efficient but less creative?
- Is knowledge flowing across departments or getting stuck?
- Do employees feel they are learning and exposed to fresh ideas, or just reinforcing what they already know?
"An AI business case must include 'people' KPIs such as employee satisfaction and not just financial outcomes," added Orsi.
The path forward requires "escape hatches” including transparency controls, exploration zones and manual overrides, which give "employees visibility into how their feeds are shaped and the option to step outside the algorithm," said Abshar.
Webinar
Sep
25
CMS Briefing: A Live Look at What’s Next in AI-Driven Platforms
Learn how leading organizations are using AI‑driven tools to publish faster, personalize smarter and stay secure.
Register
Webinar
On demand
Ready or Not: How Data-First Organizations Are Unlocking Agentforce Potential
Learn how to cut through the noise, activate Agentforce and build a Salesforce AI strategy that actually delivers.
Watch Now
Webinar
On demand
AI in Customer Service: Faster Resolutions, Happier Customers
Don’t let rising demand burn out your team. See how to build a smarter, more resilient support org.
Watch Now
Webinar
On demand
From Hype to High-Impact CX Strategies That Actually Scale
Turn buzzworthy AI and outsourcing trends into measurable CX wins with fresh 2025 data.
Watch Now
Webinar
On demand
Insights to Action Rethinking the Contact Center for Real Business Impact
Join our exclusive webinar to hear CX executives share their innovative strategies for transforming service delivery.
Watch Now
Webinar
On demand
Accelerating Healthcare Ops with AI
Sign up to see how your peers are putting AI to work — and making it stick.
Watch Now
Organizations must master a fundamental tension: how to achieve the efficiency gains of personalization while preserving the randomness that drives innovation.
As AI in the workplace continues to evolve, the stakes of getting this balance right are becoming clearer. "Over-personalization does not just hide information, it hides opportunity," Absher said.
Organizations that master this balance will harness AI's efficiency gains while preserving the cross-pollination essential for innovation. Those that do not risk creating digitally isolated workforces where short-term productivity gains come at the cost of long-term competitive advantage, cultural cohesion and the very serendipity that makes work and organizations dynamic.
Editor's Note: Read more about the two sides of AI productivity:
- Is Generative AI Actually Freeing Workers From Low-Level Work? — Take AI evangelists' promises with a grain of salt. But that doesn't mean AI success is out of reach, just that you might have to switch goals.
- The Double-Edged Sword of Using AI in Meetings — While AI can improve efficiency and inclusivity in meetings, it must be wielded responsibly. Some considerations to keep in mind.
- OpenAI Makes a Play for the Enterprise — With Connectors, Record Mode and collaboration capabilities reportedly on the way, OpenAI is positioning ChatGPT as the productivity layer for the enterprise.