The Signal Inside JPMorgan’s Monthly Update

Every month, JPMorgan quietly updates its list of highest-conviction stock picks. Most of the time, the changes are incremental — a defensive swap here, a sector rotation there. The April 2026 update is different. The bank added two software and cybersecurity names to its roster and kept a nuclear power company at the center of its energy thesis. Together, those three moves tell a coherent and important story: the AI trade is no longer about who makes the chips. It is now about who secures the infrastructure, who builds the software pipelines, and who keeps the lights on.

Palo Alto Networks (PANW): Cybersecurity as a Non-Negotiable

JPMorgan’s decision to add Palo Alto Networks to its April top picks is a statement about the maturation of AI risk. As generative AI lowers the cost and complexity of launching sophisticated cyberattacks, enterprise security budgets are not being trimmed — they are being reinforced. Cybersecurity has become, in the words of analysts tracking the space, a non-negotiable line item tied to operational and reputational risk.

The bank’s stance on PANW has been building. JPMorgan analyst Brian Essex recently noted that the selloff in security software “has moved beyond what fundamentals justify, particularly for AI beneficiaries.” The April addition formalizes that conviction. The bank is now described as “incrementally more positive” on the company as it continues to gain market share and reposition itself as a direct AI beneficiary.

• Revenue growth: 16% year-over-year to $2.47B in fiscal Q1 2026, hitting the top of guidance
• Next-gen security ARR: Up 29% to $5.85B; SASE growth of 34% to over $1.3B
• Operating margins: Exceeded 30% for two consecutive quarters
• Analyst consensus: 44 of 56 covering analysts rate PANW a Strong Buy or Buy (LSEG); consensus target implies ~31.5% upside
• CEO insider buy: Nikesh Arora made his first open-market share purchase since 2019 — JPMorgan called it a “substantial vote of confidence”

PANW’s platform thesis is straightforward: as hackers adopt AI to automate and accelerate attacks, enterprises must respond with AI-integrated defense. Palo Alto’s “Precision AI” architecture — spanning next-gen firewalls, the Cortex endpoint suite, and Prisma Cloud — positions the company as a consolidator across security operations, SASE, and cloud security. Free cash flow margins are expected to trend toward 40% over time. This is not a speculative AI play. It is a defensive infrastructure investment with expanding margins and durable secular demand.

JFrog (FROG): The Plumbing of the AI Pipeline

JFrog is the less obvious pick — and likely the more revealing one. The DevOps software platform is down roughly 23% year-to-date, and some of that underperformance has been attributed to what analysts describe as the “fear of long-term AI disruption.” JPMorgan is betting that fear is misplaced.

The bank’s thesis is that JFrog is “well positioned to benefit from meaningful AI-related tailwinds” — specifically the transition from traditional DevOps workflows to MLOps, the operational framework required to build, deploy, and manage machine learning models at enterprise scale. Every AI application needs a software supply chain. JFrog manages that supply chain. As AI development moves from experimentation into production, the demand for reliable, secure, and auditable ML pipelines will expand materially.

• Year-to-date performance: Down ~23% — creating a potential valuation entry point
• Analyst consensus: All but one analyst rates FROG a Strong Buy or Buy (LSEG)
• Consensus upside: Average analyst target implies approximately 42% upside from current levels
• Core positioning: DevOps-to-MLOps transition; binary repository management; software artifact security

The contrarian element is worth underlining. JPMorgan is adding a stock that is already down significantly on the year and doing so in a market still nervous about AI software disruption. That is a high-conviction move. The market is pricing in disruption risk; the bank is pricing in adoption acceleration. If AI model deployment in enterprise settings scales as expected, JFrog is the picks-and-shovels name for the MLOps era.

Constellation Energy (CEG): The Nuclear Safe Haven

Constellation Energy is not a new JPMorgan name — but its continued presence on the top picks list carries its own message. In an environment where oil has crossed $100 per barrel and energy market volatility is elevated, CEG functions as a nuclear safe haven: a power producer whose output is contracted, carbon-free, and structurally disconnected from commodity price swings.

The anchor of CEG’s thesis is its landmark deal with Meta. Constellation and Meta signed a 20-year power purchase agreement for the full 1,121 megawatts of output from the Clinton Clean Energy Center in Illinois. The agreement, which begins in June 2027, secures the plant’s continued operation and relicensing for two decades — replacing the state’s zero-emission credit program with a direct market-based contract. Meta’s rationale was explicit: the company stated that nuclear power is critical to sustaining its artificial intelligence ambitions and clean energy goals.

• Meta PPA: 20-year agreement for 1,121 MW of emissions-free nuclear output; commences June 2027
• Capacity: 55 GW combined after closing the Calpine acquisition in January 2026; nation’s largest private-sector power producer
• Nuclear fleet capacity factor: 94.7% for full year 2025, reaching 96.8% in Q3 2025
• Earnings growth guidance: 13%+ adjusted operating earnings growth through 2030
• Capital return: $5 billion buyback program; dividends growing 10% annually
• Expansion optionality: Evaluating advanced reactor and SMR deployment at Clinton site

The stock has pulled back significantly from its October 2024 peak — trading near $275 versus highs above $400 — as guidance for 2026 came in technically in-line but failed to clear an exceptionally high bar set by the market’s AI-utility enthusiasm. That reset may be the opportunity. Management’s projection of 20%+ base EPS growth through 2029 suggests the current friction is a timing issue, not a structural problem. The Meta deal begins flowing revenue in 2027. The Calpine integration is expected to be more than 20% accretive to adjusted EPS in 2026. The nuclear fleet does not fluctuate with natural gas prices.

The Macro Frame: Hardware to Software to Infrastructure

Step back from the individual names and a coherent rotation is visible. The first phase of the AI trade rewarded semiconductor and data center hardware companies. That phase is maturing. Capital is now moving toward the software layer — who secures it (PANW), who builds and manages it (FROG) — and toward the physical infrastructure layer that powers it (CEG). These are not speculative bets. They are investments in the operational requirements of an AI economy that is moving from proof-of-concept into deployment at scale.

JPMorgan’s April reshuffle is also geopolitically calibrated. The removal of EQT from the list — replaced in part by energy names more insulated from natural gas volatility — reflects the bank’s read on the current U.S.-Iran conflict backdrop and $100+ oil. Nuclear, by contrast, is structurally immune to that dynamic. Security software spending, similarly, tends to be maintained even as companies cut discretionary tech budgets. JFrog’s case rests on enterprise adoption timelines — the one variable that carries execution risk.

What Investors Should Watch

• PANW: Next earnings report for NGS ARR trajectory and platformization metrics; any acceleration in enterprise consolidation wins
• FROG: Enterprise customer additions; any evidence of MLOps pipeline adoption in Fortune 500 deployments; management commentary on AI-driven demand
• CEG: Crane Clean Energy Center reconnection timeline update from PJM (currently pushed to 2031); formal commencement of Meta PPA in 2027; SMR/advanced reactor permitting progress at Clinton
• Macro: Direction of oil prices and their effect on energy market sentiment; enterprise IT budget decisions in Q2 2026; AI capex guidance from hyperscalers

Bottom Line

JPMorgan’s April update is not noise. Adding PANW and FROG while maintaining CEG is a structured, thematic statement: the bank believes the next leg of AI-driven returns will accrue to the companies that protect, operate, and power the AI stack — not just the companies that built the initial compute layer. Each of these names carries a different risk profile. PANW has margin expansion and a consolidation thesis behind it. FROG has a valuation reset and a contrarian adoption argument. CEG has long-duration contracted cash flows and a nuclear scarcity premium that no commodity shock can replicate. Together, they represent a rotation worth tracking closely.

For informational purposes only.