Oracle, HPE, and Adobe Earnings: Testing the Enterprise AI Spending Thesis

Now I have comprehensive research. Let me compile the report.

1. 🔑 Key Points

• The infrastructure-first thesis remains intact, but cracks are appearing: Oracle's $523 billion backlog and 40-44% cloud growth guidance, HPE's $6.8 billion AI server order book, and Adobe's AI-influenced ARR surpassing $5 billion collectively validate that enterprise AI spending is real — yet all three stocks are down 12-23% YTD, signaling markets increasingly doubt the conversion of investment into profitable returns amid $700 billion in hyperscaler capex and rising corporate debt levels.

• Tariff headwinds create a two-speed AI economy: While Gartner projects $2.52 trillion in global AI spending in 2026 (up 44% YoY), U.S. tariffs have driven IDC to halve its broader IT spending growth forecast. This bifurcation means AI infrastructure spending is accelerating even as traditional enterprise IT budgets face compression — benefiting Oracle and HPE's AI-specific businesses while pressuring Adobe's software seat-growth narrative.

• The "Trough of Disillusionment" thesis is the real test this earnings week: Gartner has placed AI in its Trough of Disillusionment throughout 2026, predicting that most AI will be sold by incumbent software vendors during renewal cycles rather than through new project spend. This week's earnings will reveal whether Oracle can convert its record backlog, whether HPE's back-half-loaded AI server revenue is materializing, and whether Adobe's credit-based monetization can prove the software layer of AI generates genuine revenue uplift.

---

2. The Earnings Week Setup: Three Companies, One Critical Question

This section examines the specific earnings catalysts for Oracle, HPE, and Adobe, and why their collective results matter far beyond individual stock reactions.

• All three companies report during a single week (HPE on March 9, Oracle on March 10, Adobe on March 12), creating a rare sequential read across the AI spending stack from infrastructure to software.
• Each company enters earnings in a significant drawdown: Oracle is down over 23% YTD, Adobe down ~19%, and HPE down ~12%, despite strong underlying AI demand metrics.
• The convergence of these reports will serve as the first major data point of the 2026 Q1 earnings season for the enterprise AI thesis.

HP Enterprise (HPE) is down 12% to $21.13 despite 150% Networking growth, net income fell to $305M from $1.37B; Oracle (ORCL) is down 21% to $152.96 with 68% IaaS growth and a $523B backlog; Adobe (ADBE) is down 19% to $283.62, with $10.03B cash flow. These drawdowns are remarkable given the strength of the underlying AI demand metrics. The market is sending a clear signal: it no longer wants to hear about backlogs, pipelines, and future revenue potential. It wants to see profitable conversion — now.

Earnings reports next week test whether HP Enterprise's Juniper integration works, Oracle's cloud revenue matches its backlog, and Adobe's AI features drive growth. This framing captures the essential question at each layer of the enterprise AI stack. The collective outcome will either re-energize or further erode the narrative that AI spending is a self-sustaining supercycle.

YTD Stock Performance vs. Revenue Growth Expectations (March 2026)

---

3. Oracle: The Cloud Infrastructure Pivot Under Stress

Oracle's Q3 FY2026 results on March 10 will test whether the company can convert its record-shattering backlog into billable revenue while managing an increasingly precarious balance sheet.

• Oracle has guided for cloud revenue growth of 40-44% in USD, with analysts expecting quarterly cloud revenues to surge 43% to approximately $8.9 billion — representing about 52% of total group sales.
• The company's remaining performance obligations (RPO) of $523.3 billion (up 433% YoY) signal massive forward demand, but total debt has climbed 40% to $124 billion, and free cash flow has turned deeply negative.
• The Oracle-OpenAI Stargate project complications — with the Abilene expansion cancellation just days before earnings — add fresh uncertainty to Oracle's AI infrastructure narrative.

3.1 The Backlog-to-Revenue Conversion Challenge

Management guided for cloud revenue growth of 40-44% in USD, backed by a record $523.3 billion remaining performance obligations. These are staggering numbers that would be the envy of any enterprise technology company. Yet the market's skepticism is rooted in a fundamental question: how quickly can Oracle convert this backlog into billable revenue?

Analysts expect quarterly cloud revenues to surge 43% to $8.9 billion, representing about 52% of total group sales. This represents a dramatic shift — from a decade ago, when Oracle was primarily known for its on-premise database business. The re-rating from legacy software vendor to AI infrastructure provider is the core investment thesis, but it requires sustained execution.

The quarter under review reflected AI-driven demand for Oracle Cloud Infrastructure, which expanded 68% year over year in second-quarter fiscal 2026. The growth trajectory is genuinely impressive, and if Oracle can maintain this pace, the thesis holds. However, while cloud revenue is accelerating, infrastructure buildout typically compresses margins in the near term.

3.2 The Balance Sheet Concern

The most material risk to Oracle is not demand — it's financing. Legacy software revenues declined 3% in the fiscal second quarter, and free cash flow turned deeply negative at $10 billion to $13.2 billion while capex reached $12 billion. Total debt climbed 40% to $124 billion, amplifying leverage concerns.

In early February 2026, Oracle also filed a $25 billion bond offering and a $20 billion equity distribution agreement to fund its data center buildout. This financing package, while necessary, is the kind of capital structure decision that spooks fixed-income and equity investors simultaneously. Morgan Stanley analyst Keith Weiss warned that Oracle's AI expansion leaves "little room for error," projecting cumulative capex of $275 billion through the end of its fiscal 2028.

3.3 The Stargate Complication

Just two days before Oracle's earnings, Oracle and OpenAI cancelled plans to expand their flagship AI data center campus in Abilene, Texas, after lengthy negotiations broke down over financing arrangements and OpenAI's changing capacity projections. While the overall collaboration between Oracle and OpenAI continues — last July, Oracle agreed to develop 4.5 gigawatts of data center capacity for OpenAI, and that agreement is still progressing as planned — the optics heading into earnings are poor.

My take: Oracle's Q3 report will likely show strong cloud growth in the 40%+ range, meeting or slightly exceeding guidance. But the critical watchpoints are: (1) the pace of RPO-to-revenue conversion, (2) capex guidance for Q4 and beyond, and (3) management's commentary on the Stargate situation. Any sign of customer concentration risk or financing pressure will accelerate the sell-off. The stock's 52-week range of $118.86 to $345.72 tells you everything about the market's ambivalence.

---

4. HPE: The Infrastructure Layer's Ground Truth

HPE's Q1 FY2026 results on March 9 provide the most direct read on whether enterprise and sovereign AI infrastructure demand is translating into actual hardware shipments.

• HPE guided for Q1 revenue of $9.0-$9.4 billion, with Wall Street expecting ~$9.3 billion (roughly 19% YoY growth), though the quarter faces sequential decline pressure from AI server shipment timing.
• The company booked $6.8 billion in AI system orders in FY2025 with more than 60% from sovereign and enterprise customers, but revenue recognition has been persistently back-end weighted.
• HPE's transformation into a "networking-centric company" via the Juniper acquisition creates both opportunity and integration risk.

4.1 AI Server Revenue Timing

HPE's AI systems revenues in the first quarter of fiscal 2026 are likely to remain uneven due to the lumpy nature of large sovereign AI deals. The company expects AI server revenues to decline sequentially as several large AI projects are expected to ship later in fiscal 2026.

This is the frustrating reality of the AI infrastructure buildout at the enterprise and sovereign tier. Unlike hyperscalers who consume compute capacity continuously, HPE booked $6.8 billion in new AI system orders in FY 2025, including approximately $2.0 billion in Q4, with more than 60% of cumulative AI orders from sovereign and enterprise customers. But management cited elongated conversion cycles due to procurement, data center readiness, and power/cooling constraints, especially for large sovereign deals.

4.2 The Juniper Integration and Networking Pivot

HPE's most recent quarter showed 150% Networking revenue growth tied directly to the Juniper deal, but net income collapsed to $305 million from $1.37 billion a year earlier, weighed down by integration costs and a $1.621 billion goodwill impairment charge in its Hybrid Cloud segment.

The networking revenue surge is impressive, but the profit impact of integration is significant. The combined networking business delivered $2.8 billion in Q4 FY 2025 revenue (+150% YoY) with 23% operating margin, supporting the company's higher-margin mix. The critical question for Q1 is whether HPE can demonstrate improving margins as integration progresses, while simultaneously proving that AI server order momentum hasn't stalled.

4.3 Tariff Exposure

Of the three companies, HPE is the most directly exposed to tariff pressures. HPE continues to navigate a complex macro environment, including tariffs, trade policy changes and geopolitical uncertainty, which is likely to have negatively impacted overall financial performance in the fiscal first quarter. As management looks to 2026, they will draw on supply chain expertise to secure critical commodity supply and exercise pricing management discipline. They expect DRAM and NAND costs to continue to increase in 2026, the majority of which they expect to pass to the market while monitoring demand.

My take: HPE's earnings will likely show a mixed picture — solid networking revenue growth masking continued AI server revenue lumpiness. The stock trades at just 7.9X forward earnings, making expectations relatively de-risked. The key signal for the broader AI thesis is whether HPE's AI server pipeline (which management says is "multiples of backlog") is converting into new orders at a sustained pace. If not, it suggests the enterprise AI infrastructure buildout is slower than the hyperscaler-driven narrative implies.

---

5. Adobe: The Software Monetization Litmus Test

Adobe's Q1 FY2026 results on March 12 represent the single most important data point for whether AI software monetization is actually working at scale.

• Adobe guided for Q1 revenue of $6.25-$6.30 billion (~10% YoY growth) with non-GAAP EPS of $5.85-$5.90, and Wall Street expects $6.28 billion in revenue and $5.87 in EPS.
• The company's FY2026 ARR growth target of 10.2% is anchored by Firefly's credit-based monetization, tier migration, and AI-influenced ARR that already exceeds one-third of its total book of business.
• Adobe faces an existential question from investors: is generative AI a growth accelerator or an existential competitive threat?

5.1 The Credit-Based Monetization Flywheel

Adobe's Firefly monetization strategy is arguably the most sophisticated AI revenue model in enterprise software. Management reported a 3x QoQ increase in generation, with rising consumption driven by more apps enabled (Photoshop, Lightroom, Premiere, Firefly), expanding media types (notably video), broader workflows (ideation with Firefly Boards, bulk actions), and improving model output quality.

As usage increases, higher-resolution and denser outputs consume more credits, nudging users toward premium tiers and add-ons. This usage-to-upgrade pattern, combined with migration to Creative Cloud Pro tiers, provides a durable path to ARPU expansion. This is the critical mechanism — if Adobe can demonstrate accelerating credit consumption and tier upgrades in Q1, it validates the thesis that AI can enhance rather than cannibalize incumbent software revenue.

5.2 The Competitive Threat Debate

Ahead of Q1 FY26 earnings, Wall Street is divided on Adobe stock. While bulls are confident about the company's ability to deliver continued growth and see AI as a catalyst, not a threat, other analysts are concerned about the impact of competition from AI.

On January 5, 2026, Jefferies moved Adobe from "Buy" to "Hold," citing a "limited AI inflection" in the company's near-term revenue projections. This downgrade triggered the stock's sell-off. The key debate is whether Adobe's 10% revenue growth adequately reflects the potential of an AI-enhanced product suite, or whether competitors like Midjourney, Canva, and OpenAI's own creative tools are capping Adobe's upside.

5.3 What Q1 Results Need to Show

Investors will focus on total annual recurring revenue (ARR) and AI-driven ARR metrics. Specifically, Barclays analyst Saket Kalia expects $460 million in Q1 net new ARR and expects the company to beat his estimate, given tiering contribution and growing generative credit usage.

Moving forward, the market will be watching for two key metrics: the conversion rate of free AI users to premium tiers and the growth of the "Digital Experience" segment as enterprises look to automate their entire marketing stacks.

My take: Adobe's Q1 is a make-or-break moment for the AI software monetization narrative. The company must demonstrate that the 3x QoQ credit consumption growth from Q4 FY2025 is sustaining and, more importantly, converting to incremental ARR. If Adobe delivers $460M+ in net new ARR with visible AI-influenced acceleration, it will be the strongest evidence yet that the software layer of AI spending is real. A miss or even an in-line result with cautious commentary could extend the "Software Slaughter" narrative that has plagued the sector since January.

---

6. The Macro Backdrop: Tariffs as a Selective Headwind

Tariff pressures are creating a two-speed technology economy that differentially impacts infrastructure versus software spending.

• U.S. tariffs have fundamentally altered IT spending forecasts, with IDC halving its IT spending growth projection from 10% to 5% in the wake of trade policy changes.
• Hardware-dependent companies like HPE face direct cost inflation on imported components, while cloud-first businesses like Oracle and software companies like Adobe experience more indirect effects.
• Despite tariff headwinds, Gartner projects global IT spending to reach $6.15 trillion in 2026 (up 10.8%), with AI infrastructure driving the lion's share of growth.

6.1 The Tariff Impact Matrix

Technology leaders face four key areas of exposure: Higher direct costs: New tariffs could raise prices on imported hardware, infrastructure and other tech components. Supply chain disruptions: If US inventories run low, companies may face delays or struggle to find new suppliers. Rising indirect costs: Cloud, software or IT service vendors may increase prices to offset their own rising costs.

The impact is uneven across our three companies:

While semiconductors may receive some exemption protection, most enterprise IT hardware — AI servers, routers and storage units — will bear the full tariff burden. This makes HPE's margin commentary particularly important: can the company pass through costs without destroying demand?

6.2 The Paradox: AI Spending vs. IT Budget Compression

The most fascinating dynamic in 2026 is the divergence between AI-specific spending (surging) and overall IT budgets (under pressure). Another key driver of this resilience is the continued surge in investment in the information technology sector — especially in artificial intelligence. But based on the current — but fluid — tariff schedule, IDC halved its forecast for projected IT spending growth from 10% to 5% in 2025.

The "losers" in this environment are traditional enterprise software firms that have failed to integrate generative AI into their core products. These "legacy" tech companies are finding their budgets cannibalized as CIOs divert funds toward AI infrastructure. This is the spending reallocation dynamic that makes this earnings week so consequential — it's not just about total spending, but about where the dollars flow.

---

7. The AI Spending Cycle Thesis: Validation or Challenge?

This section evaluates whether the collective earnings from Oracle, HPE, and Adobe will validate the infrastructure-to-software AI spending supercycle or reveal structural weaknesses in the thesis.

• Gartner projects worldwide AI spending of $2.52 trillion in 2026 (up 44% YoY), with AI infrastructure accounting for over $1.3 trillion — more than half of total AI spending.
• The five largest US cloud and AI infrastructure providers have committed to spending $660-$690 billion on capex in 2026, nearly doubling 2025 levels.
• Yet AI sits in Gartner's "Trough of Disillusionment," where scaling depends on demonstrable ROI rather than speculative potential.

7.1 The Infrastructure Layer: Still Accelerating

The five largest US cloud and AI infrastructure providers — Microsoft, Alphabet, Amazon, Meta, and Oracle — have collectively committed to spending between $660 billion and $690 billion on capital expenditure in 2026, nearly doubling 2025 levels.

Spending on data center systems will jump 31.7% to top $650 billion, up from nearly $500 billion in 2025, a whopping $150 billion increase in a single year. Server spending alone will rocket up 36.9% year-over-year, driven almost entirely by AI-optimized hardware.

Oracle and HPE both sit directly in the path of this spending wave. Oracle as a cloud infrastructure provider receiving hyperscaler-grade demand; HPE as a hardware vendor serving sovereign and enterprise AI buildouts. If both companies deliver strong numbers, it validates that infrastructure spending is not merely a hyperscaler phenomenon — it's permeating deeper into the enterprise.

Gartner 2026 AI Spending Distribution ($2.52 Trillion Total)

7.2 The Software Layer: The Missing Link

This is where the thesis gets tenuous. Gartner's spending on AI-optimized servers is projected to grow by 49% and to account for 17% of total AI spending in 2026. That tells you where the procurement gravity sits: compute first, then software and services aligned to that compute.

After two years of intense focus on generative AI tools, companies are redirecting spending away from experimental software layers and toward the infrastructure required to make AI usable at scale. This is potentially devastating for the software monetization narrative. If enterprises are buying infrastructure first and software second, Adobe's 10% revenue growth may actually be optimistic relative to the broader software environment.

Gartner trimmed the software growth forecast from 15.2% to 14.7%. That sounds like bad news. It's not. The revised 14.7% reflects a more sustainable growth trajectory, not a softening market. Still, the gap between 37% server growth and 14.7% software growth illustrates where the investment dollars are gravitating.

7.3 The Trough of Disillusionment Dynamic

"Because AI is in the Trough of Disillusionment throughout 2026, it will most often be sold to enterprises by their incumbent software provider rather than bought as part of a new moonshot project," said Lovelock. "The improved predictability of ROI must occur before AI can truly be scaled up by the enterprise."

This is actually the most bullish framing for Adobe and Oracle as incumbent vendors. If AI is sold through incumbent renewal cycles, companies like Adobe (which owns the creative workflow) and Oracle (which owns enterprise databases and ERP) have structural advantages. The question is whether the Trough manifests as slower adoption (bearish for near-term revenue) or consolidated spending with incumbents (bullish for market share).

"Budgets will increase for a narrow set of AI products that clearly deliver results and will decline sharply for everything else," Biederman said. This bifurcation dynamic is the single most important theme for understanding what happens this earnings week.

---

8. What Each Company Needs to Prove

A framework for evaluating the earnings results against the AI spending cycle thesis.

• Oracle needs to demonstrate accelerating RPO-to-revenue conversion and provide credible commentary on capex sustainability post-Stargate complications.
• HPE needs to show that AI server orders are translating into actual shipments, with margin improvement in the networking business proving Juniper integration value.
• Adobe needs to prove that AI credit consumption is driving incremental ARR growth beyond what simple price increases and seat additions would generate.

8.1 The Scorecard

8.2 Scenario Analysis

Scenario 1: Full Validation (30% probability) All three companies beat estimates and guide higher. Oracle shows 44%+ cloud growth with improving cash flow trajectory. HPE demonstrates AI server revenue recovery with strong new orders. Adobe reports $460M+ net new ARR with accelerating credit consumption. Market interpretation: the AI infrastructure-to-software spending cycle is intact and accelerating. Tech stocks rally broadly.

Scenario 2: Infrastructure Validates, Software Disappoints (40% probability) Oracle and HPE deliver solid results, proving infrastructure demand remains robust. Adobe meets expectations but fails to demonstrate meaningful AI-driven ARR acceleration. Market interpretation: AI spending is real but concentrated in infrastructure. The "picks and shovels" trade continues; software multiples compress further.

Scenario 3: Mixed Signals Across the Board (25% probability) Each company delivers on some metrics while missing others. Oracle hits cloud revenue but guides cautiously on capex. HPE shows order growth but revenue timing remains lumpy. Adobe beats on revenue but misses on net new ARR. Market interpretation: the AI spending cycle is maturing and becoming more unpredictable. Selectivity increases.

Scenario 4: Broad Disappointment (5% probability) All three companies miss or guide lower, citing tariff uncertainty, customer budget freezes, or execution challenges. Market interpretation: the AI spending cycle has peaked. Significant sector-wide de-rating.

---

9. The Debt-Fueled Infrastructure Question

The most underappreciated risk to the AI spending cycle thesis is the growing reliance on debt to finance infrastructure buildouts.

• Oracle's total debt has climbed to $124 billion (up 40% YoY), with the company needing to raise $50 billion through debt and equity to fund its data center expansion.
• Hyperscaler capex of $660-$690 billion in 2026 is increasingly funded through long-term debt instruments, including Alphabet's unprecedented 100-year bond offering.
• The IMF has flagged that "higher leverage could amplify shocks if returns fail to materialize" — a direct warning about the sustainability of AI infrastructure spending.

But as the expansion accelerates, debt financing is becoming more prevalent, increasing leverage. This shift introduces notable risks: higher leverage could amplify shocks if returns fail to materialize, or if broader financial conditions tighten, adversely impacting firms and raising concerns about spillovers to the broader financial system.

On Feb. 10, 2026, Alphabet issued $20 billion in bonds to finance AI infrastructure, including a 100-year offering that represents the company's longest-dated debt issuance. Alphabet's move is just the latest in a growing trend as tech giants turn to long-term debt to smooth massive upfront costs and preserve balance sheet flexibility.

Oracle is the most exposed company in this earnings trio. Morgan Stanley projects cumulative capex of $275 billion through the end of its fiscal 2028. At that pace, total debt could exceed $400 billion. The key question investors should focus on: is Oracle building infrastructure that will generate returns commensurate with the debt being taken on? The Stargate cancellation raises doubts about the reliability of its largest customer commitments.

Oracle Financial Trajectory: Revenue Growth vs. Debt Expansion

---

10. Broader Market Implications and Investment Positioning

This section examines how the earnings outcomes will ripple through the broader technology sector and investor decision-making.

• The earnings trio will heavily influence sentiment for the broader Q1 2026 earnings season, particularly for AI-exposed stocks across infrastructure, software, and semiconductor sectors.
• Enterprise AI spending concentration with incumbent vendors favors Oracle and Adobe but creates winner-take-most dynamics that may not lift all boats.
• The risk-reward profile differs dramatically: Oracle offers the highest upside but with the greatest execution and balance sheet risk; HPE offers the safest valuation floor; Adobe offers the most direct read on software monetization.

10.1 The Spending Cycle Hierarchy

Based on the research, the enterprise AI spending cycle operates in a clear hierarchy:

1. Semiconductor layer (Nvidia, AMD) — validated and accelerating
2. Infrastructure/cloud layer (Oracle OCI, AWS, Azure, GCP) — validated but under profitability scrutiny
3. Hardware/server layer (HPE, Dell, SMCI) — validated but lumpy and margin-challenged
4. Software application layer (Adobe, Microsoft, Salesforce) — partially validated, monetization still proving out
5. AI-native startup layer (OpenAI, Anthropic, Cohere) — growing but far from self-sustaining

The entire cohort of pure-play AI vendors — including Cohere ($150 million ARR), Mistral (~$400 million), Perplexity ($148 million annualized), and others — likely accounts for less than $35 billion in projected combined 2026 revenue. This gap between infrastructure investment and software/application revenue is the central tension in the AI spending cycle. This week's earnings will tell us whether that gap is narrowing or widening.

10.2 Investment Implications

For investors, the positioning matrix looks like this:

If you believe the full cycle validates:

• Oracle offers the most leveraged upside, with analyst price targets implying 70-85% upside from current levels, but requires conviction on balance sheet sustainability
• Adobe offers the safest entry point for AI software monetization, trading at compressed multiples with de-risked expectations

If you believe infrastructure validates but software lags:

• HPE at 7.9X forward earnings offers the cheapest exposure to AI infrastructure with Juniper as a margin expansion catalyst
• Avoid or reduce positions in high-multiple software names until ARR acceleration is demonstrated

If you believe the cycle is peaking:

• The debt burden across Oracle, hyperscalers, and AI startups creates systemic risk. Defensive positioning with cash or value-oriented tech names is warranted

10.3 The Opinionated View

My assessment is that Scenario 2 (Infrastructure Validates, Software Disappoints) is the most likely outcome this week. The infrastructure spending cycle has too much momentum — $700 billion in hyperscaler capex, sovereign AI programs, and enterprise buildouts — to stall in Q1 2026. Oracle will likely deliver strong cloud numbers and HPE will show continued AI order momentum.

But Adobe's Q1 is where the thesis will face its real test. The 10% revenue growth guidance, while solid, is not the "AI inflection" the market needs to see. The credit consumption data will be interesting but is still too early in its lifecycle to move the needle on overall revenue growth. The market wants to see net new ARR acceleration that breaks Adobe out of its pre-AI growth trajectory — and that may not happen this quarter.

The deeper problem is the Trough of Disillusionment dynamic. AI investments will increasingly be driven by organizational maturity and proven business value, rather than visionary pilot projects. Gartner said AI has fallen into the "trough of disillusionment," the penultimate phase of its hype cycle before reaching the "plateau of productivity." This means enterprise software spending on AI will be slow, deliberate, and tied to contract renewals — not the explosive growth trajectory that justified elevated software multiples.

The AI spending cycle is real, but it is an infrastructure-first cycle that may take 2-3 more years to fully cascade into software revenue acceleration. This week's earnings will likely confirm that reality, offering validation for the infrastructure layer while leaving the software monetization question unanswered.

---

11. 📚 Recommended Topics for Further Exploration

• Hyperscaler Capex ROI Timelines: A deep analysis of when $700 billion in AI infrastructure spending will generate sufficient cloud revenue to justify the investment, and what failure scenarios look like.

• Sovereign AI Infrastructure Programs: How European ($200 billion AI Continent Action Plan), Japanese (¥1 trillion annually), and Middle Eastern sovereign AI initiatives are creating a second wave of demand beyond US hyperscalers — and which vendors benefit most.

• AI Software Monetization Models Compared: A comparative study of credit-based (Adobe), consumption-based (Oracle), and seat-based (Microsoft Copilot) AI monetization approaches and which will ultimately prove most effective.

• The Enterprise AI Trough of Disillusionment: What historical hype cycle troughs (cloud 2012-2014, mobile 2009-2011) tell us about the likely duration, depth, and eventual recovery of enterprise AI spending.

• Debt-Financed AI Infrastructure and Systemic Risk: An examination of parallels between today's AI capex boom and historical infrastructure investment bubbles (telecom in 1999-2001, railroads in the 1880s), and whether the leverage building in Oracle, SoftBank, and others poses broader financial stability risks.