AI Adoption Outpaces Trust, Microsoft Sets NTLM Deadline, Right to Repair Expands

The episode centers on the expanding adoption of artificial intelligence (AI) tools among workers alongside a notable decline in confidence. According to a Manpower Group study cited by Dave Sobel, AI confidence among workers decreased by 18% even as usage increased by 13% over the past year. This divergence highlights a governance and operational gap for MSPs, as enterprise clients confront both the potential and the risks of AI-enabled solutions, facing unresolved issues of output reliability, oversight, and liability when missteps occur.

Supporting this trend, findings from the Stanford University Institute for Human-Centered Artificial Intelligence indicate that nearly 30% of AI chatbot users encountered harmful suggestions. While these statistics lack detailed breakdowns – such as which platforms or definitions of “harmful” – they shape widespread client perceptions and intensify scrutiny of AI guidance provided by IT service providers. Meanwhile, enterprise vendors like Zendesk report improved satisfaction rates from automated resolutions but emphasize the costly need to overhaul workflows and data management to effectively harness AI benefits.

Additional focus is given to Microsoft’s scheduled deprecation of the NTLM authentication protocol, replaced by newer mechanisms that are not yet fully deployed or reliable. Dave Sobel notes that legacy systems depending on NTLM present tangible operational and legal risks for MSPs, as clients may face authentication failures or re-enable insecure protocols unless thoroughly audited. Elsewhere, the "right to repair" movement is gaining ground as the Environmental Protection Agency affirms farmers’ rights to repair their own equipment, with broader implications for IT hardware access and vendor-dependent service models.

The confluence of these developments underscores the importance for MSPs and IT leaders to shift focus from product access and resale toward risk governance, lifecycle planning, and documenting client decisions—especially in AI, authentication methodologies, and hardware maintenance. Mitigating liability, clarifying accountability with clients, and tracking evolving vendor and regulatory actions are essential to maintain relevance and safeguard operations as service and product access models change.

Three things to know today

00:00 Workers Use More AI But Trust It Less, Creating New Service Risks

03:44 Microsoft Plans NTLM Phase-Out Despite Unfinished Kerberos Replacement Technology

06:32 Google, Adobe Launch AI Subscriptions While OpenAI Retires GPT-4o Next Month

10:52 EPA Ruling Lets Farmers Repair Equipment, Pressures Tech Right-to-Repair Laws

This is the Business of Tech.

Supported by: