Search
SW News

Java 27 vs Java 25: 9 JEPs Add Post-Quantum Defense [2026]

Yusuf Demir
Yusuf DemirCloud & Software Reporter
12 min read
Java 27 vs Java 25: 9 JEPs Add Post-Quantum Defense [2026]

Oracle shipped Java 27 on September 15, 2026, and the release is forcing a question that enterprise engineering teams have been putting off for months: is it time to move past Java 25, the platform’s current long-term support branch? The headline feature is post-quantum cryptography support, a defensive upgrade aimed at protocols that intelligence agencies and security researchers warn could be broken once quantum computers mature. But the Java 27 vs Java 25 debate goes well beyond cryptography. It touches virtual threads, desktop tooling, and the broader direction Oracle is taking a language that still anchors a huge share of enterprise software, cloud backends, and financial systems.

This piece breaks down what actually shipped, how Java 27 compares against Java 25 feature by feature, where it sits against Kotlin, C#, and Go, and why the decision to upgrade is more complicated than a simple version bump. We also look at the surrounding developer-tools ecosystem, including Microsoft’s CMake Tools 1.24 update for C++ teams, because Java rarely lives in isolation inside a modern polyglot stack.

Google · Preferred Sources

Don't miss new tech stories on Google

Add TrendinTech once in the Google app and our stories appear in your news suggestions.

Add Now

What Oracle Shipped in Java 27

Oracle’s September 15 announcement frames Java 27 as the newest release of its application development platform, continuing the six-month release cadence the company adopted after Java 9. Counting preview features, the release bundles nine JDK Enhancement Proposals (JEPs), a number that places it in line with recent feature releases rather than a stripped-down interim drop. The centerpiece is strengthened support for post-quantum cryptography, which Oracle positions as a forward-looking defense against decryption techniques that would become viable with sufficiently powerful quantum hardware.

Java 27 is not a long-term support release. That distinction still belongs to Java 25, which shipped in 2025 and will keep receiving Oracle’s extended support commitments long after Java 27’s standard support window closes. That asymmetry is exactly why the Java 27 vs Java 25 comparison matters: most production teams are deciding whether to track Oracle’s latest feature work or stay anchored to the LTS branch until the next long-term release arrives.

Post-Quantum Cryptography: Why It’s the Headline Feature

Post-quantum cryptography refers to encryption algorithms designed to resist attacks from quantum computers, which could theoretically break widely used public-key schemes like RSA and elliptic-curve cryptography. Oracle’s framing of Java 27 leans heavily into this angle, strengthening the platform’s cryptographic libraries so that applications built on Java can adopt quantum-resistant algorithms without ripping out their existing security stack.

For banks, insurers, and government contractors running Java in production, this is the kind of change that shows up on compliance roadmaps years before it becomes a regulatory requirement. Security teams tend to move early on cryptographic migrations precisely because swapping out encryption schemes across a sprawling codebase takes far longer than patching a library version. The practical effect is that Java 27’s post-quantum work matters more to architecture and security teams than to day-to-day application developers, at least in this release cycle.

Virtual Threads Get Sharper in Java 27

Virtual threads, introduced as a finalized feature in Java 21, remain one of the platform’s biggest concurrency shifts in two decades, letting developers write blocking-style code that scales to millions of lightweight threads instead of being capped by the operating system’s thread limits. Java 27 continues refining that groundwork, and it shows up most clearly in Helidon 27, Oracle’s cloud-native framework for Java microservices, which is built around a fast web core designed specifically to take advantage of virtual threads.

Teams already running high-throughput services on Java 21 or Java 25 will recognize the programming model immediately. The code below reflects the kind of structured concurrency pattern that virtual threads enable, the same general approach Helidon 27 builds its request-handling core around.

try (var executor = Executors.newVirtualThreadPerTaskExecutor()) {
    List<Future<String>> results = requests.stream()
        .map(request -> executor.submit(() -> handle(request)))
        .toList();

    for (Future<String> result : results) {
        System.out.println(result.get());
    }
}

The pattern itself is not new to Java 27, but the framework-level support around it is. Helidon 27 and the refined runtime behavior in Java 27 push more of the thread-management complexity down into the platform, which is the general direction Oracle has been taking Java’s concurrency story since virtual threads first landed.

Java 27 vs Java 25: Feature-by-Feature Comparison

The table below lays out how Java 27 stacks up against Java 25 on the points that matter most to teams deciding whether to upgrade now or wait for the next LTS release.

CategoryJava 25 (LTS)Java 27
Release typeLong-term supportFeature release (non-LTS)
Release date2025September 15, 2026
Oracle support commitmentExtended, multi-yearStandard six-month cycle
Post-quantum cryptographyBaseline supportStrengthened support, Oracle’s lead feature
Virtual threadsFinalized (since Java 21)Continued refinement, deeper Helidon integration
Companion cloud frameworkHelidon on prior release lineHelidon 27, fast web core built for virtual threads
Desktop and client toolingJavaFX on prior release lineJavaFX 27, refreshed for rich client applications
JDK Enhancement ProposalsVaries by prior cycleNine JEPs, including preview features
Recommended for production LTS trackingYesOnly for teams tracking feature releases

Read as a whole, the table shows why Java 27 vs Java 25 is less a story of one release beating another and more a story of two different adoption strategies. Java 25 is the safe, supported baseline. Java 27 is where Oracle is testing the features that will likely get folded into the next LTS release down the line.

Helidon 27 and the Cloud-Native Push

Helidon 27 is Oracle’s open-source, cloud-native framework for building Java microservices, and its release alongside Java 27 is not a coincidence. The framework’s fast web core is designed around virtual threads from the ground up, which means services built on Helidon 27 can handle large volumes of concurrent requests without the thread-pool tuning headaches that have plagued Java web applications for years.

For teams already running containerized Java workloads, Helidon 27 is worth evaluating alongside other lightweight frameworks in the space. It slots into the same conversation that cloud infrastructure teams have been having about container orchestration costs, a topic covered in depth in our ECS vs EKS cost comparison, since the framework you pick for your services often determines how much orchestration overhead you end up paying for.

JavaFX 27 Brings Life Back to Java Desktop Apps

JavaFX 27 extends the Java 27 release story beyond the server room. It is a framework for building rich, interactive client applications, and its presence alongside a cryptography-and-concurrency-focused platform release signals that Oracle has not abandoned desktop Java, even as most of the ecosystem’s attention has shifted to cloud services and AI-assisted coding tools. JavaFX remains the framework of choice for a meaningful slice of enterprise internal tooling, trading desks, and scientific computing interfaces that never needed a browser front end in the first place.

Developers picking a client framework today are not just choosing between JavaFX and the usual suspects. Many are weighing Java entirely against JavaScript-based desktop frameworks, a decision that increasingly loops back into the same editor and tooling choices covered in our breakdown of Zed vs VS Code vs JetBrains Air, since IDE support for a given framework often tips the scale as much as the framework’s technical merits.

Should Enterprises Upgrade From Java 25 LTS?

The honest answer for most organizations is no, not yet, and not because Java 27 lacks merit. JetBrains’ September coverage of the release noted that JDK 27 arrives with preview features developers generally do not need to rewrite applications for unless they are actively using those specific previews. That is a deliberately low-urgency framing, and it matches how most production teams should treat a non-LTS release: evaluate it, test the post-quantum cryptography changes in a staging environment, and hold off on a full production migration until the features prove stable across a couple of release cycles.

The calculation changes for teams building net-new services or running Helidon-based microservices where the virtual-thread improvements translate directly into infrastructure savings. The table below breaks down the upgrade decision by team profile.

Team profileRecommended pathWhy
Regulated finance or healthcare backendStay on Java 25 LTS, pilot post-quantum changes in stagingCompliance teams need a long support window and audited stability
New cloud-native microservicesEvaluate Java 27 with Helidon 27Virtual-thread integration can cut infrastructure costs on high-concurrency services
Desktop or internal tooling teamsTest JavaFX 27 in a sandboxRefreshed client framework, low risk since these apps are rarely internet-facing
Security and cryptography teamsUpgrade testing environments nowPost-quantum migrations take years to plan across a large codebase
Small teams without dedicated DevOpsWait for the next LTS releaseNon-LTS releases add tracking overhead without a long support guarantee

How Java 27 Stacks Up Against Kotlin, C#, and Go

Java’s competitive position has not changed dramatically with this release, but it is worth placing Java 27 in context against the languages that regularly show up in the same hiring requisitions and architecture reviews. Kotlin, which runs on the same JVM and interoperates directly with Java libraries, continues to benefit from anything Oracle does to the underlying platform, including the virtual-thread and post-quantum work in Java 27. Teams running Kotlin services get these improvements essentially for free once they target the newer JDK.

C# and the .NET runtime remain Java’s closest structural competitor for enterprise backend work, and Microsoft’s own developer-tools push, including the CMake Tools 1.24 update for C++ teams released in September, shows that the broader ecosystem is moving on a similar cadence of incremental, reliability-focused releases rather than dramatic rewrites. Go, by contrast, continues to compete less on language features and more on simplicity and deployment footprint, an argument that has not been touched by anything in Java 27.

The net effect is that Java 27 does not change the calculus for teams choosing a language from scratch. It reinforces Java’s position for teams already invested in the JVM ecosystem, particularly those who also lean on AI-assisted coding tools, a workflow shift we examined in our look at how AI coding agents affect developer output, since language choice increasingly intersects with which tools can reliably generate and review code in that language.

The Historical Arc: From Java 8 to Java 27

Context helps explain why Java 27 looks the way it does. Java 8, released in 2014, became the longest-running default version in enterprise history, with many shops only migrating years later. Oracle’s shift to a six-month release cadence after Java 9 was meant to break that inertia, pairing frequent feature releases with LTS checkpoints every two years, a pattern that produced Java 11 in 2018, Java 17 in 2021, Java 21 in 2023, and Java 25 in 2025.

Java 27, arriving as the second feature release after Java 25, fits neatly into that rhythm. It is the kind of release that previews what eventually lands in the next LTS version, which is why its post-quantum cryptography work matters even to teams that have no intention of running Java 27 itself in production. History suggests many of these features will reappear, hardened and finalized, when Oracle ships its next long-term release.

CMake Tools 1.24 and the Cross-Language Build Picture

Microsoft’s CMake Tools 1.24, reported in September 2026, focuses on reliability fixes and productivity improvements for C++ developers rather than headline features. It is a smaller story than Java 27, but it matters to the same audience: engineering organizations running polyglot stacks where Java services call into native libraries, or where build systems have to coordinate Java, C++, and other native code in the same pipeline.

That kind of cross-language coordination is exactly where developer tooling choices start to matter as much as language features. Teams managing these build pipelines alongside containerized deployments often end up making the same infrastructure decisions covered in our guide to setting up cloud-based coding agents, since automating repetitive build and test tasks across multiple languages is one of the clearer wins AI tooling offers right now.

The Broader Developer-Tools Ecosystem in October 2026

Java 27 did not ship into a vacuum. The same window saw LangChain’s JavaScript and TypeScript SDK reach version 1.5.16 and Google release version 1.0 of its MCP Toolbox Java SDK, both signs that the tooling layer sitting on top of core languages is maturing just as quickly as the languages themselves. The MCP Toolbox release in particular matters to Java shops specifically, since it gives JVM-based teams a more direct path to wiring Java services into the Model Context Protocol ecosystem that AI agent frameworks increasingly rely on.

That overlap between core language releases and AI-tooling SDKs is becoming the norm rather than the exception. Developers picking an editor or assistant workflow this quarter are effectively also picking how well that workflow will track new JVM and SDK releases going forward, a dynamic we have tracked in our comparison of routing between Claude, ChatGPT, Gemini, and Copilot for day-to-day coding tasks.

Market Impact: What This Means for Enterprise IT Budgets

The market impact of Java 27 is less about licensing costs and more about planning cycles. Post-quantum cryptography migrations are multi-year projects once an organization commits to them, and security teams that start testing now, even on a non-LTS release, save themselves a scramble later when regulators or customers start requiring quantum-resistant algorithms as a contractual baseline. That is a pattern already playing out in cybersecurity more broadly, where compliance requirements tend to arrive well after the technical groundwork has quietly been laid.

On the infrastructure side, Helidon 27’s virtual-thread-first design is the more immediate budget lever. Services that previously needed large thread pools and correspondingly larger memory allocations to handle concurrent load can, in principle, run leaner once rebuilt around virtual threads. For cloud teams already comparing GPU and compute pricing across providers, the kind of infrastructure math covered in our breakdown of AI model pricing gaps is a reminder that compute efficiency gains at the runtime level compound with smarter provider selection at the infrastructure level.

Predictions: Where Java Goes From Here

  • Post-quantum cryptography support finalized in Java 27 will likely become a standard, non-optional feature by the time the next LTS release ships, following the usual pattern of preview-to-mainstream adoption.
  • Helidon 27’s virtual-thread-first architecture will push more Java shops to evaluate Helidon as a lighter alternative to heavier legacy frameworks over the next two release cycles.
  • JavaFX’s modest revival in Java 27 is unlikely to reverse the long-term decline of native Java desktop development, but it will keep the framework viable for the enterprise internal-tooling niche where it already dominates.
  • Expect closer integration between JVM tooling and AI agent protocols, building on the momentum of Google’s MCP Toolbox Java SDK, as more Java teams wire their services directly into agent-based development workflows.
  • The real adoption wave for Java 27’s features will arrive with the next LTS release, not with Java 27 itself, consistent with how Java 21’s virtual threads only saw broad production use once Java 25 arrived as a stable LTS target.

What Developers Should Do This Quarter

For most teams, the practical move is to treat Java 27 as a preview of what is coming rather than a mandate to migrate immediately. Spin up a test environment, run your existing suite against the new post-quantum cryptography defaults, and check whether any of your dependencies break under the refined virtual-thread behavior. If you are building something new, especially a high-concurrency service, Helidon 27 is worth a serious look before you default to whatever framework your team used two years ago. And if your organization leans on AI-assisted coding, it is worth checking how your editor of choice handles the newest JDK before committing, a question we address directly in our rundown of the best AI code editors for 2026.

Frequently Asked Questions

Is Java 27 a long-term support release?
No. Java 27 is a standard six-month feature release. Java 25, released in 2025, remains the current long-term support version and will keep receiving Oracle’s extended support commitments.

What is the biggest new feature in Java 27?
Strengthened post-quantum cryptography support is the feature Oracle leads with in its September 15, 2026 announcement, alongside continued refinement of the virtual-threads concurrency model.

Should I upgrade production systems from Java 25 to Java 27 right away?
Most teams should not. JetBrains’ September coverage of the release noted that the preview features in JDK 27 do not require rewriting existing applications unless you are actively using those specific previews. Testing in staging is reasonable; a full production migration is not urgent for most organizations.

What is Helidon 27?
Helidon 27 is Oracle’s open-source, cloud-native framework for building Java microservices. It is built around a fast web core designed specifically to take advantage of Java’s virtual threads.

Does Java 27 affect JavaFX desktop development?
Yes. JavaFX 27 shipped alongside Java 27 as a refreshed framework for building rich, interactive client applications, extending the release beyond server-side use cases.

How many JDK Enhancement Proposals are in Java 27?
Nine, when preview features are counted, according to coverage of the release.

How does Java 27 compare to Kotlin and C# this release cycle?
Kotlin runs on the same JVM and inherits Java 27’s virtual-thread and cryptography improvements automatically once a project targets the newer JDK. C# and .NET are moving on a similarly incremental cadence, reflected in Microsoft’s own September 2026 CMake Tools 1.24 update for C++ developers, rather than shipping a comparably large feature jump.

Where does Java 27 fit in Java’s broader release history?
It is the second feature release after the Java 25 LTS, following Oracle’s six-month cadence that has produced LTS checkpoints roughly every two years since Java 11 in 2018, through Java 17 in 2021, Java 21 in 2023, and Java 25 in 2025.

Related Coverage

Yusuf Demir

Yusuf Demir

Cloud & Software Reporter

Yusuf Demir is the Cloud & Software Reporter at TrendinTech, where he covers cloud infrastructure, enterprise platforms, developer tools and the digital transformation of businesses in the UK and the United States. He is most interested in the gap between vendor roadmaps and the systems engineers actually run, and in what the cloud bill looks like once the free credits expire.

All stories by Yusuf Demir (355)

Related Articles