Posts

The Jellyfish AI Engineering Trends study of 7,548 engineers found a stark pattern: the heaviest AI token users produced twice the PR throughput but consumed ten times the token budget. More tokens do not equal more productivity — they equal a steeper cost curve that most engineering leaders aren’t measuring. What Is the Jellyfish AI Engineering Benchmark — and Why Should You Care? The Jellyfish AI Engineering Benchmark is the largest continuous dataset of real-world AI coding behavior ever assembled: as of early 2026 it covers 1,000+ companies, 200,000 engineers, and 37 million pull requests analyzed over rolling quarters. Unlike survey-based studies that capture developer sentiment, Jellyfish pulls instrumented telemetry — actual PRs merged, code churn rates, token consumption logs, and review cycles — making it a ground-truth view of what AI coding tools actually produce rather than what developers believe they produce. The benchmark is updated quarterly and published at jellyfish.co/ai-engineering-trends. ...

AI coding tool switching costs are higher than the monthly subscription fee suggests. The real cost includes proprietary config formats that don’t travel across tools, workflow muscle memory that takes two to four weeks to rebuild, and BYOK restrictions that may lock your agent-mode usage to a vendor’s own models. This guide breaks down every layer of cost and gives you a concrete playbook to build a portable stack. What Are AI Coding Tool Switching Costs? (Beyond the Monthly Fee) AI coding tool switching costs refer to the full set of friction and expense involved in moving from one AI-assisted development environment to another — and they go far beyond canceling a subscription and signing up for a new one. According to a 2026 Parallels survey, 94% of IT leaders now list vendor lock-in as a primary concern as AI adoption accelerates, and for good reason: the switching costs are both financial and operational. On the financial side, developers carry duplicate subscriptions for one to three months during transitions, pay for productivity dips while muscle memory rebuilds, and sometimes discover that BYOK savings evaporate once API token usage scales up. On the operational side, proprietary config files (like Cursor’s .cursorrules) must be manually rewritten, IDE keybindings must be reconfigured, and team conventions documented in one tool’s format need porting. GitHub Copilot accounts for 42% of all tool-switcher origin points in 2026, suggesting that the first migration is the most common — and the most instructive for understanding what you’re actually paying to leave behind. ...

Long-running AI coding agents use iterative execution loops where the model plans, acts, evaluates, and loops again — while single-prompt workflows send one request and stop. Choosing the wrong architecture for a task costs you hours of debugging or wasted tokens. This guide explains when each approach wins, how the top tools implement them, and what failure modes to watch for. What Is an Execution Loop? The Agentic Architecture Explained An execution loop is a software architecture where an AI agent repeatedly cycles through plan → act → observe → evaluate until a termination condition is met, rather than generating a single response and stopping. In 2026, every major AI coding tool implements some form of execution loop: Claude Code’s CLI loop with compaction, Cursor’s Agent Mode and Background Agents, Windsurf’s Cascade flow, OpenAI Codex’s three-tier hierarchy, and Gemini CLI’s continuous session. The defining characteristic is that the agent maintains state across multiple LLM calls, using the output of each step as input to the next. Gartner projects 40% of enterprise applications will embed task-specific AI agents by 2026, up from less than 5% in 2025 — and execution loop architecture is the foundation of all production-grade agentic systems. The key takeaway: execution loops are not just “longer prompts” — they are fundamentally different control flow structures that require different engineering approaches. ...

Enterprise AI coding adoption is no longer a forward-looking trend — it’s the new baseline. Over half of the Fortune 500 companies are paying for Cursor seats. GitHub Copilot has penetrated 90% of the Fortune 100. And yet the data reveals a paradox that every senior engineer and engineering leader needs to understand: 84% of developers use AI coding tools, but only 29% actually trust the output. This guide breaks down what’s happening at Fortune 500 companies, what the security and governance implications are, and what it means for developers building in enterprise environments in 2026. ...

Claude Code holds a 91% customer satisfaction score and NPS of 54 — the highest marks in the AI coding tool category as of January 2026 — while growing from 3% to 18% at-work adoption in just eight months. The satisfaction gap over GitHub Copilot (4.8/5 vs. 4.1/5) is wide enough to matter, and 46% of senior engineers now call it their most-loved tool. Here’s what the data shows and why it happened. ...

84% of developers now use AI coding tools, yet 38% of Fortune 500 companies have already experienced security incidents from those tools. This checklist gives CTOs a structured framework to evaluate AI coding assistants across six critical dimensions—security, compliance, ROI, governance, and vendor accountability—before signing any enterprise contract. Why CTOs Need a Formal AI Coding Tool Evaluation in 2026 AI coding tools have crossed from optional to essential in enterprise software development. By 2026, AI tools write 41% of all code—up from 25% in 2024—and 90% of Fortune 100 companies have deployed AI coding assistants. Yet the adoption curve has outpaced governance: only 29% of developers trust AI-generated code output, down from 40% in 2024, even as usage accelerates. This trust gap is not a sentiment problem—it reflects measurable production risk. Developers now spend 11.4 hours per week reviewing AI-generated code versus 9.8 hours writing new code, a reversal of the 2024 pattern that creates a hidden labor cost most procurement models ignore. The real stakes: 38% of Fortune 500 companies have experienced security incidents tied directly to AI coding tools. CTOs who treat AI coding tool selection as a feature-comparison exercise—rather than a governance and risk decision—are creating liability. A formal evaluation framework, not a vendor demo checklist, is the minimum responsible standard for 2026 procurement. ...

JetBrains Central is an enterprise-grade agentic platform that lets teams govern, orchestrate, and observe AI coding agents — Junie, Claude, Codex, Gemini CLI, and custom agents — from a single control plane. It launched Early Access in Q2 2026 with design partners including Google Cloud, Anthropic, and OpenAI. What Is JetBrains Central? The Agentic Platform Explained JetBrains Central is a managed infrastructure platform for agentic software development — it provides the governance layer, execution infrastructure, and semantic context that enterprise teams need to run AI coding agents reliably at scale. Unlike individual AI coding tools (Copilot, Cursor, Junie standalone), JetBrains Central is not an IDE plugin or a chat assistant. It is the control plane that sits above all those tools and coordinates their work across your development organization. Think of it as a Kubernetes for AI coding agents: it schedules workloads, enforces access policies, tracks costs to teams and projects, and surfaces logs so you know exactly what every agent did and why. The platform launched in Early Access on March 24, 2026, with design partners already including Google Cloud, Anthropic, and OpenAI — a signal that JetBrains is not building in isolation but is deeply integrated into the major AI provider ecosystems. For teams currently evaluating agentic engineering, JetBrains Central is the only solution in the JetBrains ecosystem that provides organization-level visibility into agent activity rather than per-developer fragmentation. ...

AI code security scanning tools in 2026 have become non-negotiable for any team shipping software at scale. With 45% of AI-generated code introducing OWASP Top 10 vulnerabilities and 93% of organizations using AI-generated code without applying the same security standards as traditional code, the right scanner can be the difference between a secure release and a headline breach. This guide compares Snyk, Checkmarx One, Veracode, and Black Duck across SAST, SCA, DAST, AI-specific detection, pricing, and real-world fit. ...

The EU AI Act imposes legally binding obligations on developers and deployers of AI systems in the EU, with the primary enforcement deadline of August 2, 2026. However, the AI Omnibus deal reached in May 2026 significantly changed which requirements apply on that date — extending certain Annex III high-risk AI system deadlines to December 2027. This guide tells you exactly what still hits in August 2026, what got delayed, and the specific technical steps engineering teams must take now. ...

Enterprise AI coding shadow IT is the fastest-growing governance blind spot in software development today. According to Menlo Security’s 2025 report, 57% of employees using free-tier AI tools input sensitive company data — and 68% access these tools through personal accounts, completely bypassing enterprise security controls. This isn’t a minor policy gap. It’s a systemic exposure that’s costing organizations millions and creating direct regulatory liability. The Shadow AI Coding Crisis: What the 57% Statistic Really Means Enterprise AI coding shadow IT refers to the unauthorized use of AI-powered coding assistants, autocomplete tools, and generative code platforms by developers who bypass official IT procurement and approval processes. The 57% figure from Menlo Security’s 2025 research doesn’t measure accidental misuse — it measures developers deliberately routing sensitive source code, internal APIs, and business logic through personal-account AI tools to avoid corporate oversight. A companion stat makes the picture worse: Awareways 2025 found that 73% of employees use AI tools their organization has not approved, and Lenovo’s April 2026 research found 70% of enterprise AI now operates entirely outside IT oversight. The average enterprise has 14 distinct AI tools in active use, but IT is aware of only 4–5 of them (Enterprise AI governance industry analysis 2026). Shadow AI isn’t a fringe behavior — it’s the default behavior. The 57% figure is a floor, not a ceiling, and for development teams specifically, the exposure is deeper because the data at risk isn’t just business communications: it’s proprietary source code, architectural diagrams, authentication logic, and database schemas. ...