The Big Con of Agentic AI

2026-07-10 · Source: Towards Data Science · Field: Business & Management — Corporate Strategy & Leadership, Consulting & Professional Services, Operations & Process Management · Depth: Intermediate, extended

Summary

The article "The Big Con of Agentic AI" argues that increasing reliance on agentic AI mirrors the structural risks observed in the management consulting industry, leading to a gradual forfeiture of human agency and critical capabilities. It draws parallels to Mariana Mazzucato and Rosie Collington's 2023 book *The Big Con*, detailing how individuals, organizations, and societies risk ceding cognition, judgment, and accountability. Specific concerns include cognitive debt, loss of institutional knowledge, and unexpected financial burdens, citing Microsoft's 2026 internal AI coding license cuts and Salesforce's projected \$300 million payment to Anthropic. Geopolitical risks are also highlighted, exemplified by the June 2026 US government order for Anthropic to suspend access to Fable 5 and Mythos 5 for foreign nationals. The piece advocates for a deliberate "third path" to reclaim agency, balancing AI's benefits with active governance.

Key takeaway

For Directors of AI/ML evaluating adoption strategies, recognize that uncritical integration of agentic AI risks hollowing out internal expertise and creating costly vendor dependencies. You must actively invest in building your team's cognitive capacity and institutional memory, diversify AI suppliers, and ensure human oversight truly challenges AI outputs. Prioritize structural sovereignty to prevent long-term atrophy and maintain control over critical operations.

Key insights

Over-reliance on agentic AI structurally erodes human agency and critical capabilities, mirroring the "con" of management consulting.

Principles

• Repeated delegation fosters "unlearning by not doing."
• Outsourcing strategic functions erodes core identity.
• Tools must serve people through deliberate choice.

Method

Adopt a "third path" by maintaining individual cognitive struggle, investing in organizational institutional memory and diversified AI, and implementing public policies for structural sovereignty and accountable human-in-the-loop systems.

In practice

• Modify AI outputs to retain cognitive agency.
• Diversify AI suppliers for critical functions.
• Invest in open-weight models for public services.

Topics

• Agentic AI
• Organizational Dependency
• Cognitive Atrophy
• AI Governance
• Geopolitical Risk
• Institutional Knowledge

Best for: Investor, CTO, VP of Engineering/Data, Director of AI/ML, Consultant, Policy Maker

Related on AIssential

• Agentic AI: What Leaders Wish They Knew Sooner — Effective agentic AI integration requires strategic process redesign, gradual trust-building, and human adaptation to new collaborative roles.
• Senior execs worry about AI system lock-in — AI adoption is creating significant vendor lock-in and dependency issues for enterprises, demanding strategic control and oversight.
• Weekly Top Picks #115 — AI's rapid evolution presents both significant opportunities and complex challenges across various societal and economic domains.
• The Next Category Is Not Agentic AI — The next enterprise AI category is Sovereign Intelligence Architecture, not agent-only propositions, focusing on integrated human-digital workforce governance.
• Why Over 40% of Agentic AI Projects May Fail — Agentic AI project success hinges on organizational readiness and progressive autonomy, not just model capability.
• Your AI strategy has a trust problem, not a tooling problem — Organizational trust and employee agency are critical for AI innovation, outweighing tooling advancements.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Towards Data Science.