Editorial
Research Report
The AI Transformation 100: Concrete Ideas for Fixing How We Work
Early 2026 operations revealed a system defined not by its resilience, but by its capacity for automated violence. During the critical opening hours of Operation Epic Fury, US forces struck thousands of targets across Iran based on recommendations generated through the Maven Smart System (MSS) and its commercial large language model, Claude, which was integrated into Palantir’s proprietary platform.
This exercise in algorithmically driven warfare coincided with the release of the Office of Management and Budget’s Memorandum M-25-21. This directive explicitly instructs federal agencies to accelerate artificial intelligence adoption, remove bureaucratic barriers and “lean forward on adopting effective, mission-enabling AI.”
Table of Contents
- What Is Memorandum M-25-21?
- The Collapse of the Workforce Pipeline
- The Human & Environmental Toll of Generative Models
- A Permanent Transfer of Control
- Automated Governance in Civilian Systems
- The Loss of Human Agency & Looming Leadership Vacuum
What Is Memorandum M-25-21?
Memorandum M-25-21 instructs agencies to build and retain AI-ready workforces while sharing code, models and data across the government, effectively amplifying the message that machine assistance is no longer optional; it is mandatory.
High-impact AI tools are positioned as public goods whose rapid deployment must outweigh concerns about performance or interpretability. Consequently, structural incentives have aligned: national security requirements demand ever-faster targeting cycles, while civilian directives encourage automation to maximize efficiency.
The narrative celebrates this speed, yet it masks how these same mandates erode human oversight. When strategic guidance frames human review as a bottleneck and encourages the delegation of risk acceptance to subordinate officials, the stage is set for the total replacement of deliberative judgment with algorithmic throughput.
The structural reveal is unavoidable: the state does not merely adopt AI to enhance public services; it legislates a surrender of human discretion to automated systems.
Related Article: The Conquest of Meaning: AI Procurement and the Erosion of Command
The Collapse of the Workforce Pipeline
The justification for mass automation is almost always framed through the lens of financial prudence. Memorandum M-25-21 specifies that agencies must “maximize the value of existing investments” and reuse AI models and data to avoid duplication, a policy that implicitly rewards projects that deliver more outputs with fewer personnel.
Private enterprises are following an identical logic. A joint report from the Brookings Institution and Opportunity@Work, as examined in a recent Fast Company analysis, noted that over 15 million US workers without four-year degrees hold jobs highly exposed to AI. Of these individuals, approximately 11 million occupy "gateway positions" that serve as essential entry points into higher-paying careers.
Gateway roles build foundational competencies in clerical and administrative tasks, yet 62% of workers in these positions lack college degrees. When organizations automate these functions to improve their revenue-per-employee metrics, they effectively eliminate the very environments that produce long-term skill development. The US Government Accountability Office (GAO) found that staffing reductions at the Internal Revenue Service in 2025 led to the loss of sixty-three employees specifically working on AI initiatives, noting that the IRS lacked a formal plan to identify and address these critical skills gaps.
Executives are beginning to speak openly about these consequences: LHH research on the c-suite shows that executive turnover fell from 43% to 19%, yet nearly half of these leaders now cite AI and emerging technologies as their top development priority. Human expertise is becoming a scarce commodity at the exact moment that strategic clarity and disciplined decision-making are most needed. Financial efficiency is being achieved not through innovation, but through the deletion of apprenticeships.
The structural irony surfaces: each automated process that raises margins simultaneously destroys the pipeline of future talent, ensuring that organizations cannot sustain or even interpret the very systems they deploy.
The Human & Environmental Toll of Generative Models
Evidence of skill atrophy is no longer merely anecdotal. A GAO assessment on generative AI emphasized that the massive equipment powering these models consumes substantial amounts of water and electricity, yet companies rarely disclose their full resource usage. US data center electricity demand, which was already 4% of national consumption in 2022, is on track to reach 6% by the end of 2026.
These energy demands reflect a runaway complexity rather than any deliberate human improvement. While generative systems risk replacing workers or producing dangerous deepfakes, the GAO assessment noted that definitive statements about their societal effects remain impossible due to corporate secrecy and rapid evolution.
This technology’s opacity exacerbates skill decay: when an output arrives instantly and the explanation vanishes behind a proprietary model, humans learn to accept the result rather than understand the process. Healthcare experiments illustrate this paradox clearly. A 2026 pilot program in Utah allows an autonomous AI agent to renew prescriptions for 192 chronic disease drugs. Although safeguards require human physicians to review the AI’s decisions for the first 1,250 patients and mandate an escalation protocol for complex cases, early commentary noted that accountability for errors remains dangerously unclear. By removing routine interactions between patients and physicians, the program risks reducing the quality of preventive care. The pattern repeats across sectors: generative systems promise efficiency yet degrade human skills and oversight while consuming natural resources at an unsustainable scale.
The structural reveal here is that the purportedly “intelligent” system is cannibalizing both the environment and the human expertise necessary to supervise it.
A Permanent Transfer of Control
Defense procurement demonstrates how modern contractual structures institutionalize a state of permanent dependency. Palantir’s Maven Smart System evolved from a modest $91.2 million contract in 2020 into the central node of all US military targeting. A five-year, indefinite-delivery contract worth $480 million was awarded in May 2024, followed by a $99.8 million task order to expand access across the services. By May 2025, the Department of Defense raised the contract ceiling to $1.3 billion, and in July 2025, it consolidated roughly seventy-five existing contracts into a massive $10 billion enterprise agreement.
The Palantir platform integrates sensor feeds, command-and-control workflows and machine learning modules into a proprietary ontology. Once combatant commands ingest their data and train their personnel on this specific system, substitution becomes prohibitively disruptive. The algorithmic network thus becomes the “platform of platforms,” where switching costs embed institutional memory into commercial software.
The program’s expansion reveals the erosion of human oversight: Operation Epic Fury compressed the targeting cycle to produce one thousand recommendations per hour, leaving minimal time for substantive human review. As the tempo approaches cognitive limits, the human decision-maker becomes little more than a rubber stamp, formalizing the algorithm’s selection. As the Electronic Frontier Foundation (EFF) noted, this shows that privacy protections currently rest on executive discretion rather than law.
The structural reveal identifies a permanent transfer of control: the state’s lethal authority now depends on private platforms whose incentives may diverge from the public interest, eroding democratic accountability.
Related Article: The Persuasion Paradigm: When Security Hands Over the Keys
Automated Governance in Civilian Systems
Civilian administration is mirroring this military consolidation. The Centers for Medicare & Medicaid Services launched the Wasteful and Inappropriate Service Reduction (WISeR) model in January 2026. This voluntary six-year program leverages AI and machine learning, combined with human clinical review, to determine Medicare payments for select services in six states. Participants receive financial incentives tied to the value of care deemed "unnecessary," effectively empowering private companies to streamline prior authorization and lower spending.
While WISeR frames AI as a way to protect taxpayers, it also introduces a dynamic where algorithmic determinations influence patient access to life-altering treatments. Companies apply proprietary models to decide whether procedures like knee arthroscopy or electrical nerve stimulators are medically necessary. Although recommendations for non-payment must be validated by licensed clinicians, the system’s financial structure heavily incentivizes denials.
Simultaneously, the Electronic Frontier Foundation reported that CMS has refused to release details of WISeR under the Freedom of Information Act, using AI to evaluate medical necessity without any public transparency.
The structural irony is evident: a model designed to reduce wasteful services introduces an opaque algorithmic gatekeeper, delegating discretion over patient care to private vendors paid based on the savings they generate. This mechanism centralizes authority within AI systems while reducing the role of physicians and patients in determining appropriate care.
Webinar
Jun
3
From AI Investment to CX Results: What Enterprise Leaders Need to Know
Move beyond experiments. See how top enterprises scale AI for CX results.
Register
Webinar
Jun
4
Content Strategy Leaders Live: Managing Risk, Compliance & AI in Financial Services
Learn how financial services leaders are modernizing content systems without disrupting trust, compliance or experience.
Register
Webinar
Jun
16
From Legacy to Launch-Ready: How Gainbridge Made Its Website a Marketing-Led Growth Engine
Join in to learn how a D2C annuity brand gave marketing full website ownership — without slowing down or risking compliance.
Register
Webinar
On demand
Content Strategy Leaders Live: Managing Scale, Safety & AI in Manufacturing
How manufacturing leaders update content systems to handle product complexity and scale while integrating AI safely.
Watch Now
Webinar
On demand
AI With Intent: How CX Leaders Decide Where Automation Belongs (And Where It Doesn't)
Senior CX leaders from REI & Hyatt reveal their framework for using AI to boost retention.
Watch Now
Webinar
On demand
From Content Sprawl to Competitive Advantage: A Knowledge Operations Roadmap
How automation, AI and a centralized content platform turn knowledge fragmentation into a scalable competitive advantage.
Watch Now
The Loss of Human Agency & Looming Leadership Vacuum
The combined effect of accelerated adoption, workforce elimination, skill atrophy and vendor lock-in is producing a profound leadership crisis. As gateway jobs disappear, so too do the training grounds that develop diagnostic intuition.
The GAO noted that the IRS currently lacks a plan to fill its AI skills gaps, a situation that serves as a microcosm of a much broader pattern. Executives recognize this deficiency; nearly half of the leaders surveyed in the LHH 2026 report identified AI literacy as their primary development need, even as strategic clarity and disciplined decision-making remain their top constraints.
When late-career executives delay retirement and succession pipelines remain undeveloped, organizations become dangerously brittle. A small cohort retains control, but there is no new generation prepared to assume responsibility. Meanwhile, generative AI’s environmental costs continue to rise, and policy solutions remain focused on encouraging innovation rather than reducing usage or ensuring safety. Public trust is eroding as privacy rights depend on corporate negotiation rather than statutory protection.
The structural reveal is not merely a matter of inefficiency, but one of inversion: systems designed to increase capability are instead removing human agency. Leadership once grounded in experiential knowledge is being replaced by managers who depend entirely on algorithms they cannot audit. Governance by machine promises certainty yet delivers fragility. In this vacuum, both private vendors and government agencies expand their control, cementing the final vassalage of compute over public life.
Learn how you can join our contributor community.