Business of Tech: Daily 10-Minute IT Services Insights

AI Raises Workloads and Burnout: HBR Study, Medical Risk, and New Governance for MSPs

Feb 11, 2026

Ask episode

Chapters

Transcript

Episode notes

Artificial intelligence (AI) is intensifying workloads rather than alleviating them, leading to increased burnout and declining decision quality, according to findings published in the Harvard Business Review and cited by Dave Sobel. The episode underscores that AI lowers the cost of producing outputs such as drafts and summaries but raises throughput targets and introduces new verification burdens. Economic gains from AI remain concentrated where capital and skilled labor already exist, while negative impacts—like displacement and wage pressure—are felt locally. These dynamics highlight the need for robust governance, particularly for managed service providers (MSPs) who deploy AI solutions.

Supporting studies referenced include the International AI Safety Report, which details heightened uncertainty around AI development and its risks, as well as research from Oxford documenting the unreliability of AI chatbots in real-world medical decision-making. Experts warn that rapid automation without corresponding improvements in control systems creates structural constraints, making traditional software governance frameworks inadequate for unpredictable AI behaviors. Without proactive measures, these gaps risk exacerbating economic inequality and liability in regulated environments.

Additional developments include OpenAI’s release of upgraded agent features—such as GPT-5.2, improved context retention, managed shell containers, and a new skills standard—presented as operational enhancements but raising concerns about black-box context handling, auditability, and dependency risk. T-Mobile’s AI-powered live translation service offers greater convenience but eliminates audit trails, shifting compliance risk to customers and prohibiting independent verification. Quark Cyber’s launch of an internal cyber risk score introduces further complexity, as the scoring methodology is embedded within a financial product structure and lacks transparent validation.

For MSPs and IT service leaders, the key takeaway is to treat new AI features and risk metrics as tools with significant tradeoffs. AI deployments should focus on governance layers that include workload caps, quality gates, and measurable outcomes rather than simply accelerating productivity. New features should be used for low-stakes workflows and carefully avoided in high-risk or regulated contexts unless auditable controls and deterministic checkpoints are established. Vendor-managed risk scores and warranties require independent validation before being positioned as client-facing truth standards.

Four things to know today

00:00 Harvard, Oxford Studies Find AI Raises Workload, Delivers Inadequate Medical Advice

05:01 OpenAI Updates Deep Research and Adds New Agent Runtime Capabilities

07:33 T-Mobile Tests Real-Time Call Translation Built Into Its Network

09:17 Cork Cyber Rolls Out New Risk Score for Managed Service Providers

This is the Business of Tech.

Supported by: ScalePad
Small Biz Thoughts Community