The Pendulum Swing of AI Adoption

October 2026

Author: Jim Routh

We are living in an unprecedented era of technological disruption. AI is reshaping how we live, work, and think faster than any previous wave of change. History is full of technology‑driven shifts met with pessimism, but none match the speed of AI adoption, nor have any warned of humanity’s possible demise. Historians who study the atomic bomb note that nuclear war remains a low-probability, high-impact threat. Similarly, a coordinated swarm of AI agents could sabotage critical water, energy, and air infrastructure, making the planet uninhabitable. 

OpenAI’s recent disclosure of the Hugging Face attack illustrated a swarm of > 1,000 agents working together to evade AI containment despite enterprise guardrails.  The challenge is amplified by the strong business appetite for AI: lower costs, faster time‑to‑market, and higher quality promise lucrative returns. Consequently, many enterprises are under pressure to deploy AI while simultaneously using it to detect and remediate system vulnerabilities before cyber‑attacks explode. This double‑edged race strains the professionals, employees and contractors, who must implement AI under great uncertainty.

Recent surveys show that > 50 % of workers in professional, scientific, and technical services are questioning their employment decisions . Cyber‑ and IT‑infrastructure teams are working around the clock to patch vulnerabilities discovered within minutes. Employees report reduced sleep, poor nutrition, and dwindling support for dependents, an unsustainable mix that harms health and family life.

Board members and senior leaders intensify the dilemma. They demand rapid skill development to keep pace with lightning‑fast AI advances, yet they also require immediate remediation of known vulnerabilities, often without allocating human resources.  Leaders push for aggressive AI deployment while planning staff reductions based on the anticipated benefits of an “agentic” workforce.

Historically, large‑scale technology rollouts succeeded when senior leadership explained what’s in it for the employee. When people understood the personal benefits, they embraced change.  Today, most enterprise AI initiatives are driven by aggressive milestones without a clear employee pathway. Early results show that learning efforts have limited value and business objectives remain elusive:

• 80 % of AI projects missed business goals

• 95 % of generative‑AI pilots had no P&L impact after six months

• 42 % of enterprises abandoned most AI initiatives by 2025

• Only 28% of AI use cases in IT infrastructure and operations fully met ROI expectations; 20% failed outright.

These data, combined with anecdotal evidence of employee stress, make a compelling case for a new enterprise path:

1. Identify employee value – Clarify the direct benefits each worker will gain.

2. Skill‑investment questionnaire – Ask every employee: “Which two skills will you master?”

3. Document skill definitions – Leaders coach employees to make skills granular enough for concrete development activities.

4. Provide development resources – Offer internal and external educational tools aligned with each skill.

5. Aggregate choices – HR designs a curriculum that addresses the majority of identified needs.

6. Scheduled knowledge‑share sessions – Allocate one day per week, with a 90‑minute slot for teams to discuss successes and failures, capturing feedback for continuous improvement.

Skeptics may argue there is no time for such investment. Yet enterprises currently spend far more resources hiring new AI talent than developing existing staff, even though the market is saturated with learners.  The COVID‑19 pandemic taught us to prioritize:

1. Health and well‑being – Without personal health, employees cannot support others.

2. Work‑schedule flexibility – Adjust hours to meet personal and community needs.

3. Focused work time – Align effort with critical objectives while helping colleagues when possible.

Applying these lessons to AI adoption means acknowledging that the current pace is unsustainable and must change.

Employee stress from conflicting directives is causing the loss of valuable talent  – often the very people needed after AI implementation. A longer‑term effect of rapid AI adoption is the heightened premium on uniquely human “soft” skills that AI cannot replace: stakeholder management, consensus facilitation, relationship building, coaching, mentoring, communication, aligning personal goals with business objectives, curriculum design, negotiation, and more.  As AI handles analytical tasks, these interpersonal capabilities will become even more critical. The pendulum swing to AI analytics for business objectives will likely swing back to investing in people skill development once the objective of AI mastery is realistic in future years.

Jim Routh

Jim Routh is Chief Trust Officer at Saviynt and a Fellow at the Institute for Critical Infrastructure Technology (ICIT). A board member, advisor, investor, faculty member, and mentor, he previously served in senior cybersecurity leadership roles, including as CSO/CISO for American Express, DTCC, KPMG, Aetna, CVS, and MassMutual.

About ICIT

The Institute for Critical Infrastructure Technology (ICIT) is a nonprofit, nonpartisan, 501(c)3think tank with the mission of modernizing, securing, and making resilient critical infrastructure that provides for people’s foundational needs. ICIT takes no institutional positions on policy matters. Rather than advocate, ICIT is dedicated to being a resource for the organizations and communities that share our mission. By applying a people-centric lens to critical infrastructure research and decision making, our work ensures that modernization and security investments have a lasting, positive impact on society.  

Learn more at www.icitech.org.