Tidbits from the past week…

1. CEOs report a continued increase in AI spending despite uneven and unclear ROI. Fewer than half of AI initiatives are generating positive returns today, yet nearly 70% of executives plan to increase spending in 2026. This reflects a defensive investment approach—AI-related investments are viewed as a strategic necessity, not a performance-driven decision.
2. A recently completed Brookings Nationwide AI Usage Survey found the following:
   • AI adoption is widespread personally, but shallow professionally. A majority of Americans (57%) report using generative AI for personal purposes, yet only 21% use AI in their professional roles.
   • While AI productivity gains are real, they are not broadly felt. Only 19% of respondents report any productivity improvement, and just 4% report significant gains.
   • Workers are skeptical about AI-driven job growth. One in eight respondents believes that AI will increase job opportunities in their industry over the next five years.
   • Professional AI use tends to cluster in specific activities, including document writing and editing, data analysis and forecasting, customer support and decision support.
   • Regulated sectors, e.g., healthcare, finance, and government, are adopting AI pragmatically, focusing on efficiency and support functions rather than full automation.
   • Demographic divides are material and persistent. AI use drops sharply among workers without a college education, lower-income earners, and adults over 60 years old.
3. AI is following the internet's trajectory, but faster and with higher social friction. Leaders should plan for a multi-year transition marked by volatility, role confusion, and uneven value capture rather than a clean "AI revolution" moment.
4. Trust, judgment, and human oversight remain binding constraints. Even as AI outperforms humans on standardized tests and defined benchmarks, organizations remain reluctant to cede decision authority in high-risk domains.
5. The rise of "knowledge architects," AI-assisted accountants, and evolving white-collar roles highlights a widening gap between job titles, actual work performed, and worker identity.
6. The AI "arms race" is compressing vendor differentiation. As capabilities converge, integration and ecosystem lock-ins will become the true differentiators.
7. Connected and AI-enabled consumer products, including toys, are showing systemic security and privacy weaknesses. The gap between deployment speed and safeguards is triggering bipartisan scrutiny.

1. AI agents are now outperforming human professionals in high-risk knowledge work at a fraction of the cost and time. This represents an important inflection point. It no longer merely augments expert labor; it is beginning to substitute for it in bounded domains such as arbitration, cybersecurity, and logistics.
2. Frontier models have become force multipliers. The same tools that are used for autonomous information security operations and improved vulnerability discovery lower the barrier to attack. AI is not only a productivity tool but also a systemic risk amplifier.
3. Creative labor has entered an era of synthetic substitution with the emergence of AI performers like "Tilly Norwood." Identity-based labor is now contestable! This will intensify labor disputes, redefine talent economics, and force new norms around authorship, likeness, and creative ownership.
4. Leading IP holders such as Disney are pivoting from AI resistance to rent-seeking opportunities from its properties, including AI-generated content. This signals a new equilibrium in which disruption in the creative industries will be managed through licensing and platform partnerships.
5. In the future, AI leadership will be decided by business models, not benchmarks. Technical performance rankings are no longer a reliable predictor of winners.
6. Key highlights from the recent AI league tables include:
   • Anthropic – While it flies somewhat under the radar due to its low presence in the consumer space, Claude is a go-to choice for enterprise customers.
   • Google – Recent gains were made with its release of Gemini 3 language model, but generative AI threatens to upend the search business that funds the bulk of its empire.
   • Microsoft – Its AI fortunes are tied to its relationship with OpenAI, Copilot, and its own nascent AI frontier strategy.
   • OpenAI – In many ways, it's the one to beat, especially from a consumer perspective, but it lacks a separate business to fund operations.
7. AI has extended its reach as churches across the U.S., traditionally cautious adopters of new technology, are leveraging AI to generate content, including devotional imagery, sermons, and localized messaging.

1. AI is no longer an innovation project. It is a structural force reshaping risk, labor, infrastructure, and governance. It has:
   • Introduced new systemic exposure, including model failures, data-center dependencies, AI-enabled cyberthreats, etc.
   • Reshaped the labor market by redistributing tasks across roles, amplifying a productivity divide between AI-augmented "super-users" and everyone else.
   • Transformed infrastructure requirements in areas such as computing capacity, energy availability, systems reliability, and model-training ecosystems.
   • Redefined governance with executives responsible for setting boundaries on model autonomy, ensuring "explainability" and auditability.
2. As AI shopping assistants become the primary interface for product discovery, customer acquisition is shifting from search engine optimization (SEO) to AI agent optimization (AIO).
3. Estée Lauder demonstrated the ability to translate intangible, sensory experiences into actionable commerce. Their AI Scent Advisor doubled online conversion rates versus shoppers who did not use it.
4. Generation Z (those born between 1997 and 2012) has become the first "all-AI" work cohort. 93% of Gen Z workers use two or more AI tools weekly. Their expectations are forcing employers to adapt or face talent retention issues.
5. AI is reshaping workforce structures in a slow but sweeping manner. These changes are moving in the following direction:
   • AI hits tasks before jobs
   • Since occupations are "bundles" of tasks that are not easily automated, workflows shift first, not headcount
   • New job families are emerging around AI's "backbone"
6. In the end, the most pessimistic forecast suggests 30% unemployment with double-digit GDP growth driven by automation.
7. HR has become one of the riskiest domains for AI adoption. They are heavy users of AI (65%), especially in recruiting, screening, and job description generation. Experts warn that hallucinations in evaluations could lead to discrimination claims.

1. This year has become the first true AI holiday shopping season as retailers, e.g., Amazon (Rufus) and Walmart (Sparky), rolled out shopping assistants integrated into purchasing flows. Functions include instant personalization, guided discovery, product comparisons and price checks, and "frictionless" checkout. On "Black Friday," consumers spent $11.8 billion online, with AI-driven traffic up 805% year over year.
2. As rivals vie for AI platform leadership, model leapfrogging, e.g., Google's recent Gemini upgrade, holds implications for enterprise adoption. This competitive horse race is creating pressure on organizations as they formulate AI roadmaps.
3. In spite of AI's rapid increase in usage by "individual contributors," corporate adoption is slow, uneven and unequally distributed. While skill gaps and manager hesitancy are factors, the principal bottleneck to AI adoption is the need to change workflows versus providing tools to individual workers. An explanation of the perceived slow pace may lie in the Solow Paradox. Integration takes time! Current studies suggest a lag of 2–5 years before effects are observed in productivity statistics.
4. While AI is reshaping America's $9.4 trillion labor market, traditional workforce metrics do not provide a forward-looking projection of the effects across industries. Project Iceberg, a joint-effort between researchers from MIT, National Taskforce on State AI Policy, Oak Ridge National Laboratory, and others, simulated the human-AI labor market representing 151 million workers.
5. A growing body of evidence suggests that AI "super-users," those who are most skilled in prompting, workflow design, and AI-enabled productivity, are deliberately not fully automating their own work. This behavior can be attributed to:
   • Cognitive-trust gap
   • Loss-of-agency concern
   • Professional identity protection
   • Quality control anxiety
   • Fear of skill atrophy
6. As AI reshapes the creative industries, "human-authenticity" may become a premium brand differentiator in entertainment, design, and marketing. Industry groups are preparing for dual-track workflows: AI-augmented production versus authenticity-certified human work.
7. This past week's outage, which hit CME's equity, bond and commodity futures operations, highlighted the importance of "continuity of business" planning, especially in light of enterprise dependency on hyper-scale data centers to fuel large models.

Corporate boards are divided on AI pace, governance, level of risk to accept, and ownership. Their lack of alignment may slow adoption, especially in large enterprises.

Sector Watch

• AI brokers and AI-first financial intermediaries will disrupt financial advisory, insurance brokerage, trade support, and underwriting roles. Firms that don't differentiate their service offerings through value-added human expertise will experience revenue declines and margin compression.
• As Federal, state, and local governments grapple with funding shortfalls, staff shortages, downsizing, and demand surges, deployment of "elastic" AI agents at scale signals a major turning point in public-sector automation.
• AI is institutionalizing itself within the K-12 systems. Teachers are already heavy users, FERPA-compliant tools are emerging, and students are using AI for homework help, research, scheduling, organizing notes, and receiving personalized feedback.
• Higher education signals a shift as they recognize AI as a "force multiplier." They are introducing AI-integrated degree programs across many disciplines.

Additional Insights

4. The use of generative AI is colliding with human communication norms. Issues include branding consistency, "false fluency," content overproduction, and loss of narrative judgment.
5. Corporate information security teams are faced with dramatically faster AI-enabled attack cycles. Non-technical workers may not be adequately prepared to recognize deepfake voice attacks, hyper-personalized phishing, and AI-fabricated invoices.
6. Fragmented and/or shifting Federal and state AI regulatory regimes present compliance challenges. The use of HR-related tools will likely face scrutiny for AI-mediated decisions.
7. World Models represent AI's next leap. While LLMs produce text, ideas, and answer arguments, they are not grounded in real-world physics. A world model is a representation of how the world works—physically, temporally, causally, and interactively. Future AI systems will combine an LLM frontend for language, planning, and coordination, and a world-model backend for grounded reasoning and action.

Organizational structures and spans of control are being reevaluated in situations where AI agents and humans participate in mixed workflows. Particular areas of attention include ownership, oversight, and audit rules for machine-performed work. CIOs and HR are forming joint units to manage AI deployment, training, culture change, and integration of "digital co-workers" (AI agents) into the workforce.

1. AI proficiency is becoming a condition of employment. Companies are no longer merely encouraging AI adoption—they are mandating it, with consequences for noncompliance.
2. Productivity gains are real, but AI burnout is emerging as an employee issue due to new expectations, increased speed, and the relentless pace of AI-enabled workflows.
3. While there are mixed signals concerning the impact of AI on employment, news tidbits continue to suggest growing pressure on jobs.

A more positive view…

Based on an analysis of U.S. Bureau of Labor Statistics data, higher AI usage is not strongly correlated with softer job growth.

On the flip side…

• Organizations are openly stating that AI can now perform many entry-level tasks, reducing the need for traditional early-career hires.
• AI-driven efficiencies are "wreaking havoc" on the white-collar workforce, with a focus on support functions such as HR, operations, marketing, and staff services.
• Mid-level workers displaced by AI are now applying for entry-level roles—a "double whammy" for recent graduates.
• Economic uncertainty and anticipated benefits from AI investments are contributing to hiring pauses across many sectors.

AI workforce impacts differ between large and small firms. Large organizations are reshaping organization structures and roles, while small firms are using AI defensively to remain competitive, not to shrink headcount.

1. Corporate layoffs have shifted to "AI-driven optimization," with it becoming one of the reasons to streamline white-collar staffing. Investor reaction is positive, accelerating the trend.
2. The first AI workforce transparency mandate hit congress. Senators Hawley (R-MO) and Warner (D-VA) introduced the AI Jobs Impact Clarity Act, which will require large employers and federal agencies to provide quarterly AI impact reporting in ESG-style disclosures.
3. AI is relying on a shadow labor market in which humans, e.g., Uber drivers, lawyers, doctors, musicians, etc., are being used to train and fine tune models.
4. While the use of AI is moving from experimentation to operationalization, efforts may stall as they hit a middle management chokepoint.
5. AI has entered core governance, e.g., financial reporting, signaling a shift where individuals in corporate functions become editors of AI output rather than authors.
6. Marketing and creative functions are undergoing a "quiet" disruption as they restructure their activities around human oversight, creative direction, model prompting, and rapid iteration at scale.
7. Collective robotics (drones, microbots) are paving the way for "human + swarm" operating models in which technician, dispatcher, and field-service roles shift toward oversight and exception management.
8. AI is emerging as a "front door" for non-emergency medical care. Consumers are increasingly using AI models for real-time, personalized, longitudinal health guidance.
9. While AI is replacing tasks, organizations are beginning to view entry-level hiring, especially AI-natives, as a strategic feeder for AI-augmented workflows.
10. AI-powered malware strains are adapting to defenses and being used to customize attacks. Investment will be required in threat-detection systems, adversarial testing, and staff retraining on AI-native vulnerabilities.