Don’t Let AI Slop Muck Up Your Company’s Processes

2026-06-16 · Source: Feeds - HBR.org · Field: Business & Management — Corporate Strategy & Leadership, Operations & Process Management, Human Resources & Workforce Development · Depth: Intermediate, long

Summary

Generative AI introduces a significant risk of "knowledge decay" within organizations, where the accuracy and quality of internal information deteriorate due to the widespread use of AI-generated content. This phenomenon, an organizational extension of "workslop," leads to compounding errors, eroded trust in data, and increased time spent on verification. Authors Matthias Holweg and Thomas H. Davenport, in their June 16, 2026, Harvard Business Review article, identify three core challenges: verifying AI-generated content, validating human-added value in outputs, and managing knowledge entropy, which is the gradual degradation of information as it passes through probabilistic transformer algorithms. They note that "model collapse" can occur when LLMs are trained on synthetic data. The article highlights examples in HR, scientific research, and healthcare, where AI's integration has led to issues like generic job descriptions, retracted papers, and potential malpractice concerns.

Key takeaway

For Directors of AI/ML or Operations Professionals integrating generative AI, you must proactively manage the risk of "knowledge decay" to prevent process deterioration and maintain data integrity. Implement clear strategies for tracking data provenance and restricting AI use to tasks where it genuinely adds value, rather than merely generating content. Redesign business processes to ensure AI integration enhances overall efficiency and quality, rather than introducing "workslop" that necessitates extensive human verification and erodes trust in outputs.

Key insights

Generative AI risks organizational knowledge decay through verification, validation, and entropy challenges, requiring strategic process intervention.

Principles

• AI-generated content can erode trust and quality at scale.
• Knowledge decay stems from verification, validation, and entropy.
• New technology needs process alignment for productivity gains.

Method

Organizations should track unstructured data provenance, restrict generative AI use to value-adding tasks, clearly define AI's value contribution, and assess AI's impact on entire business processes.

In practice

• Implement structured questionnaires for factual data in hiring.
• Use proprietary SLMs or customized models for insights.
• Agree on AI usage across interorganizational processes.

Topics

• Generative AI Risks
• Knowledge Management
• Process Automation
• Data Provenance
• AI Ethics
• Organizational Productivity

Best for: CTO, VP of Engineering/Data, AI Architect, Director of AI/ML, Executive, Operations Professional

Related on AIssential

• The risks of inviting AI into the heart of our economy, society and governance | Letters — The core risk of AI lies in its capacity to blur truth, erode provenance, and diminish human originality and empathy.
• Why the Model Context Protocol Proves Generative AI Engines Are Running on Empty — Generative AI's enterprise utility fundamentally depends on integration with structured, verified real-world data, not self-sufficiency.
• Companies are hyping AI the same way they talked up sustainability, but there are ways to fix that — AI washing, akin to greenwashing, misleads stakeholders by overstating AI benefits and understating risks.
• Study warns generative AI raises ML security and bias risks — Integrating generative AI into ML systems heightens security, bias, and transparency risks.
• As frontier models increasingly dictate the parameters of human discourse, clinical diagnostics, and financial risk, the lack of transparency regarding their underlying data architectures has become.. — AI faces a triadic crisis of performance degradation, environmental impact, and trust due to synthetic data and opacity.
• Don’t trust Generative AI to do your taxes — and don’t trust it with people’s lives — AI systems are currently unreliable for critical governmental functions like tax processing and military applications.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Feeds - HBR.org.