{"company":{"name":"Harness","slug":"harness","website":"https://harness.io","category":"developer-tools"},"question":"What has Harness shipped recently?","answer":"In the last 30 days, Harness shipped 36 tracked updates. The most recent was \"Harness launches Software Delivery Agent to automate governed deployments at machine speed\" on 2026-10-07.","window":{"days":30,"updateCount":36,"returned":36},"generatedAt":"2026-10-08T06:35:47.746Z","updates":[{"title":"Harness launches Software Delivery Agent to automate governed deployments at machine speed","summary":"Harness introduced the Software Delivery Agent to automate the post-code delivery process, addressing bottlenecks in review, build, test, secure, deploy, and verify stages. The agent executes controls at machine speed, providing platform and DevOps teams with a single governed path from commit to production under set policies, with built-in identity, policy, and audit controls.","date":"2026-10-07","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/resources/harness-live-demo-software-delivery-agent","publisher":"harness.io"},{"title":"Harness launches Security Testing Agent with AI-driven vulnerability prioritization and automated fixes","summary":"Harness introduced the Security Testing Agent (STA), an AI-powered tool that scans code, dependencies, and pipelines to prioritize vulnerabilities by exploitability and reachability. It generates validated fixes delivered in a single pull request, aiming to reduce manual remediation time from weeks to minutes.","date":"2026-10-06","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/resources/security-testing-agent-datasheet","publisher":"harness.io"},{"title":"Harness achieves 99.8% Security Efficacy in SecureIQLab's Cloud WAAP v5.0 validation","summary":"Harness was named a leader in SecureIQLab's Cloud WAAP v5.0 CyberRisk Validation Report, achieving 99.8% Security Efficacy and 98.2% Operational Efficiency. The report highlights Harness's 100% effectiveness against OWASP API Top 10, zero false positives, and full coverage for advanced threats including bot attacks and AI-assisted attacks.","date":"2026-10-05","dateIsEstimated":true,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/secureiqlabs-leader","publisher":"harness.io"},{"title":"Harness rebuilds AI API Classifier on Jev to cut costs and latency","summary":"Harness replaced the judgment step in its AI API classification pipeline with Jev, reducing per-endpoint classification costs by 1,200× versus the original pipeline and 150× versus a single-shot LLM. Jev’s calibrated probabilities improved accuracy and enabled a stable output contract, allowing backend changes without downstream coordination.","date":"2026-10-05","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/judgment-not-generation-rebuilding-our-ai-api-classifier-on-jev","publisher":"harness.io"},{"title":"Harness introduces knowledge graphs to unify SDLC data and power deterministic AI","summary":"Harness published a detailed guide explaining how knowledge graphs can unify fragmented software delivery lifecycle (SDLC) data across tools by establishing explicit entities, relationships, and schema-as-code contracts. The post argues this approach enables deterministic AI reasoning and operational discipline, contrasting manual data stitching with queryable knowledge systems.","date":"2026-10-05","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/knowledge-graphs-for-software-delivery-architectural-approach","publisher":"harness.io"},{"title":"Harness hosts webinar on securing AI and Agents in software delivery","summary":"Harness announced an on-demand webinar detailing its security policies, frameworks, and guardrails for protecting customer data and assets when using Harness AI and Agents. The session covers AI governance, model and data security, adversarial attack protections, third-party model usage, and secure agentic execution within pipelines.","date":"2026-10-05","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/resources/securing-ai-and-agents","publisher":"harness.io"},{"title":"Access Management","summary":"Access Management Advanced Role-Based Access Control gives granular control over what users and service accounts can do across the platform.","date":"2026-10-04","dateIsEstimated":true,"signalType":null,"signalTypeLabel":null,"sourceUrl":"https://developer.harness.io/harness-ai/use-harness-platform/platform-access-control/configure-rbac#rbac-configuration-workflow","publisher":"developer.harness.io"},{"title":"Harness explains why engineers ignore cloud cost governance and unveils Cost Management Agent","summary":"Harness argues that traditional FinOps fails engineers due to delayed feedback and tooling friction, leading to runaway cloud costs. The company introduces its Cost Management Agent to embed real-time cost governance into CI/CD workflows via automated guardrails, policy enforcement, and granular visibility across multi-cloud environments.","date":"2026-10-02","dateIsEstimated":false,"signalType":null,"signalTypeLabel":null,"sourceUrl":"https://www.harness.io/blog/why-engineers-ignore-cloud-cost-governance-and-fixes","publisher":"harness.io"},{"title":"Securing AI and Agents: How Harness Protects Your Data, Models, and Software Delivery| Harness","summary":"Harness announced a webinar detailing its security policies, frameworks, and guardrails for protecting customer data and assets when using Harness AI and agentic capabilities. The session covers model/data security, adversarial attack protections, third-party model usage, and secure agent execution within software delivery workflows.","date":"2026-10-01","dateIsEstimated":false,"signalType":"event","signalTypeLabel":"Event","sourceUrl":"https://www.harness.io/events/ai-security-governance","publisher":"harness.io"},{"title":"Harness ships 102 updates in September 2026: ChatGPT control, merge queues, and IaC secret scanning","summary":"Harness delivered 102 product updates in September 2026, including a ChatGPT plugin for pipeline management, merge queues for PRs, built-in secret scanning for IaC, and a Homebrew CLI installer. Key additions also include AI agent import for CD, API discovery-to-testing automation, and enhanced model key access for worker agents.","date":"2026-10-01","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/shipped-in-september-2026","publisher":"harness.io"},{"title":"Harness explains why engineers ignore cloud cost fixes and pushes AI-driven FinOps","summary":"Harness argues that engineers often overlook cloud cost optimization due to misaligned incentives and lack of visibility, proposing AI-powered cost management agents to embed FinOps governance into workflows. The post highlights Harness’s broader push into AI-driven cloud and AI spend management.","date":"2026-09-30","dateIsEstimated":true,"signalType":null,"signalTypeLabel":null,"sourceUrl":"https://www.harness.io/blog/why-engineers-ignore-cloud-cost-optimization-fixes","publisher":"harness.io"},{"title":"Harness launches connector for OpenAI to bring software delivery context into AI tools","summary":"Harness introduced a new connector enabling ChatGPT and Codex to inspect pipeline health, investigate build failures, and execute deployment workflows directly from AI tools. The integration aims to bridge AI-driven development with Harness's CI/CD platform.","date":"2026-09-29","dateIsEstimated":true,"signalType":"integration","signalTypeLabel":"Integration","sourceUrl":"https://www.harness.io/blog/harness-connector-for-openai","publisher":"harness.io"},{"title":"Harness report reveals AI productivity gap: 31% of dev time spent on invisible validation work","summary":"A Harness-commissioned survey of 700 developers found 31% of their time is spent on AI-related invisible validation work (e.g., reviewing AI-generated code) that current metrics fail to capture. 94% of engineering leaders admit their frameworks miss tech debt, validation time, and burnout, while 81% report increased code review time since AI adoption.","date":"2026-09-25","dateIsEstimated":true,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/state-of-engineering-excellence","publisher":"harness.io"},{"title":"Harness: AI agents deployed without adequate controls, per State of Agent DLC 2026 report","summary":"Harness found 75% of organizations have deployed AI agents in production despite lacking critical controls, with 87% experiencing agent-related security incidents and 60% overspending budgets. Confidence in testing, security, inventory, cost, and rollback hovers in the mid-70s, but actual governance tooling lags significantly.","date":"2026-09-25","dateIsEstimated":true,"signalType":null,"signalTypeLabel":null,"sourceUrl":"https://www.harness.io/state-of-agent-dlc-2026","publisher":"harness.io"},{"title":"Harness outlines 2026 cloud cost optimization strategies and AI-powered features","summary":"Harness published a comprehensive guide on cloud cost optimization for 2026, emphasizing allocation, rightsizing, idle resource management, and commitment strategies. The post highlights three new AI-powered cloud cost optimization features and introduces AI Productivity Insights and AI DLC Insights to measure ROI of AI engineering investments.","date":"2026-09-24","dateIsEstimated":true,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/cloud-cost-optimization-strategies","publisher":"harness.io"},{"title":"Harness introduces agent-native control plane for AI-driven software delivery","summary":"Harness unveiled a new control plane concept enabling AI agents to autonomously manage the full software delivery lifecycle beyond pull requests. The system uses standardized interfaces like MCP to let agents handle testing, security, deployments, rollbacks, and incident response from a single terminal. A demo showed an agent detecting a feature-flag error, requesting access, rolling back changes, verifying recovery, and documenting the incident.","date":"2026-09-24","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/resources/beyond-the-pr-the-new-control-plane-for-agentic-software","publisher":"harness.io"},{"title":"Harness publishes guide to 15 AWS cost optimization tactics","summary":"Harness released a blog post detailing 15 AWS cost optimization tactics across compute, storage, networking, and monitoring, accompanied by a quick-reference table. The post links to multiple Harness product announcements, including AI-powered cost optimization features and AI agents for software delivery.","date":"2026-09-23","dateIsEstimated":true,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/aws-cost-optimization-tactics","publisher":"harness.io"},{"title":"Harness integrates dynamic AI discovery and runtime security with Amazon Bedrock AgentCore","summary":"Harness introduced a new integration with Amazon Bedrock AgentCore, providing security teams continuous visibility and real-time threat detection across AI agent interactions on AWS. The integration enables dynamic AI discovery and runtime security for agent-based workflows.","date":"2026-09-22","dateIsEstimated":true,"signalType":"integration","signalTypeLabel":"Integration","sourceUrl":"https://www.harness.io/blog/harness-brings-dynamic-ai-discovery-and-runtime-security-to-amazon-bedrock-agentcore-gateway","publisher":"harness.io"},{"title":"Harness explains why cloud cost optimization for engineers fails","summary":"Harness argues that traditional cloud cost optimization efforts often fail engineers due to misaligned incentives and lack of visibility, advocating for developer-centric FinOps practices. The post highlights how Harness’s AI-powered tools address these gaps by integrating cost awareness into development workflows.","date":"2026-09-22","dateIsEstimated":true,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/why-cloud-cost-optimization-for-engineers-fails","publisher":"harness.io"},{"title":"Harness explores how machine learning reshapes DevSecOps and the SDLC","summary":"Harness published a blog outlining how machine learning is transforming secure software development lifecycle (SDLC) practices, including AI-assisted code review, automated triage, and new risks like model supply chain attacks. The post highlights Harness’s focus on integrating AI into DevSecOps workflows.","date":"2026-09-18","dateIsEstimated":true,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/ai-devsecops","publisher":"harness.io"},{"title":"Harness reveals enterprise AI adoption paradox in new research","summary":"Harness announced findings from a survey of 700 engineering leaders showing a confidence gap in AI agent controls despite widespread production use. The research highlights rising incidents as agents scale and outlines best practices for safe deployment.","date":"2026-09-17","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/resources/confidence-without-control-navigating-agentic-ai","publisher":"harness.io"},{"title":"Harness to host live demo of Cost Management Agent on October 14, 2026","summary":"Harness announced a live demo on October 14, 2026, showcasing its Cost Management Agent, which automates cloud and AI spend attribution, optimization, and governance. The agent uses a knowledge graph to track spend to owners and includes features like AutoStopping, Cluster Orchestration, and Commitment Orchestrator.","date":"2026-09-17","dateIsEstimated":false,"signalType":"event","signalTypeLabel":"Event","sourceUrl":"https://www.harness.io/webinars/live-demo/live-demo-cma","publisher":"harness.io"},{"title":"Harness: AI Cost Visibility — Why Policies Alone Don’t Stop Surprise Bills","summary":"Harness argues that AI spend policies alone are insufficient to prevent surprise bills, emphasizing the need for real-time cost visibility and governance. The post highlights how organizations often overlook the dynamic nature of AI costs, leading to budget overruns despite policy enforcement.","date":"2026-09-17","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/ai-cost-visibility-why-policies-alone-dont-stop-surprise-bills","publisher":"harness.io"},{"title":"Harness outlines best practices for securing Internal Developer Portals","summary":"Harness published a blog post detailing how embedding security, auditing, and governance into Internal Developer Portals (IDPs) reduces compliance risks without slowing engineering teams. The post highlights risks like infrastructure drift and workflow sprawl, emphasizing the need for centralized visibility, role-based access control, and automated audit logging to prevent misconfigurations and security incidents.","date":"2026-09-17","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/developer-portal-security-and-governance-best-practices","publisher":"harness.io"},{"title":"Beyond the prompt: governing AI behavior like you govern releases| Harness Blog","summary":"Harness unveiled AI Configs, a runtime-adjustable governance system for AI behavior that mirrors feature flag management. Teams can now swap prompts, models, or settings in production without redeploys, enforcing policies like RBAC, audit logs, and progressive rollouts. The approach targets the 51% of enterprises already running agentic AI in production but struggling with slow, linear AI deployment pipelines.","date":"2026-09-17","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/beyond-the-prompt-governing-ai-behavior-like-you-govern","publisher":"harness.io"},{"title":"Beyond the sync: Argo CD needs an Enterprise Control Plane| Harness Blog","summary":"Harness argues that Argo CD excels at Kubernetes manifest synchronization but lacks enterprise-scale workflow orchestration, governance, and AI-driven verification. The company positions its new Enterprise Control Plane as a wrapper around Argo CD to handle complex release workflows, centralized RBAC, OPA Policy as Code, and automated rollbacks via AI verification.","date":"2026-09-17","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/beyond-the-sync-argo-cd-needs-an-enterprise-control-plane","publisher":"harness.io"},{"title":"Governance is the platform problem worth solving| Harness Blog","summary":"A Harness-hosted LeadDev panel argued that traditional governance models fail for AI agents, which act faster than human review cycles. Harness and panelists from Yelp and Platformable emphasized automating governance via platform-enforced controls, audit trails, and observability for agent-driven changes.","date":"2026-09-17","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/governance-is-the-platform-problem-worth-solving","publisher":"harness.io"},{"title":"Best Kubernetes Cost Management Tools 2026| Harness Blog","summary":"Harness published a 2026 guide comparing Kubernetes cost management tools, emphasizing the need for namespace-level cost allocation and idle/unallocated cost visibility. The post positions Harness’s Cost Management Agent as a solution that integrates cost data into engineering workflows, supports AI inference workloads, and automates right-sizing and idle shutdowns.","date":"2026-09-16","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/best-kubernetes-cost-management-tools","publisher":"harness.io"},{"title":"AI Security Governance: A Practical Framework for 2026| Harness Blog","summary":"Harness published a framework for AI security governance emphasizing four pillars: policy, risk assessment, monitoring, and accountability. The post highlights regulatory pressures like the EU AI Act (2026-2027 deadlines) and the rapid adoption of AI agents in software delivery, which boards now approve without clear controls.","date":"2026-09-15","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/blog/ai-security-governance","publisher":"harness.io"},{"title":"We used our own product to cut our AI tooling bill without cutting AI usage| Harness Blog","summary":"Harness used its own Cloud Management Agent to reduce AI tooling spend without reducing usage by implementing per-developer cost visibility, weekly budgets with graduated alerts, and self-service budget increases. Cost per million tokens dropped sharply after governance launched in mid-July 2026, while total spend temporarily declined before normalizing.","date":"2026-09-15","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/we-used-our-own-product-to-cut-our-ai-tooling-bill-without-cutting-ai-usage","publisher":"harness.io"},{"title":"Policy as Code in 2026: OPA, Kyverno, Cedar and What’s Next| Harness Blog","summary":"Harness launched Policy as Code (PaC) to automate security and compliance rules as version-controlled, testable artifacts, addressing the 'Authorization Gap' in modern software engineering. The solution leverages Open Policy Agent (OPA) for general-purpose governance while integrating specialized frameworks like Kyverno for Kubernetes and Cedar for application-level authorization.","date":"2026-09-11","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/policy-as-code-in-2026-opa-kyverno-cedar-and-what-s-next","publisher":"harness.io"},{"title":"Why engineers ignore cloud costs, and how AI Cost Management Agents fix it| Harness Blog","summary":"Harness explains why engineers ignore cloud costs due to delayed feedback loops and introduces AI Cost Management Agents that provide real-time cost visibility and automated actions within engineering tools. The post highlights how these agents integrate cost governance into development workflows, including pull request comments, policy enforcement, and contextual spend tracking for AI and cloud resources.","date":"2026-09-11","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/why-engineers-ignore-cloud-costs-and-how-ai-cost-management-agents-fix-it","publisher":"harness.io"},{"title":"State of Agent DLC 2026| Harness Resource","summary":"Harness published a 2026 report revealing a 30–55 point gap between organizations' confidence in AI agent safety and their actual controls, based on a survey of 700 engineering leaders across five countries. The report highlights systemic deficiencies in testing, security, inventory, cost, and rollback readiness despite 75% of organizations running agents in production.","date":"2026-09-10","dateIsEstimated":false,"signalType":"content_marketing","signalTypeLabel":"Content","sourceUrl":"https://www.harness.io/resources/state-of-agent-dlc-2026","publisher":"harness.io"},{"title":"Organizations Are Confident Their Agents Are Behaving. But They Can't Check.| Harness Blog","summary":"A Harness blog post reveals that 700 surveyed organizations report 74-77% confidence in their AI agents' security, testing, inventory, cost, and rollback controls, but most lack dedicated tooling. Only 23% have AI-agent-specific security layers, 19% auto-block bad releases, and 44% lack active agent inventory discovery. 58% report more production incidents since deploying agents.","date":"2026-09-10","dateIsEstimated":false,"signalType":null,"signalTypeLabel":null,"sourceUrl":"https://www.harness.io/blog/the-ai-agent-confidence-gap","publisher":"harness.io"},{"title":"AI Code Review Loop in the Terminal: Introducing Harness CLI for Harness Code| Harness Blog","summary":"Harness launched CLI 3.0, unifying Harness Code with native CI/CD pipelines, AI-powered code reviews, and interactive terminal UIs. The CLI introduces a spec-driven grammar with six core verbs covering 320+ commands, enabling zero-context-switch pull request workflows, risk-bucketed AI insights, and live pipeline log streaming directly in the terminal.","date":"2026-09-09","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/ai-code-review-loop-in-the-terminal-introducing-harness-cli-for-harness-code","publisher":"harness.io"},{"title":"The Contained Pen Test: Proving Resilience Without Widening the Blast Radius| Harness Blog","summary":"Harness launched a contained pen testing feature that runs security experiments directly in CI/CD pipelines, proving system resilience under specific failure conditions without granting broad production access. Each experiment is narrowly scoped, repeatable, and produces immutable audit records.","date":"2026-09-09","dateIsEstimated":false,"signalType":"feature_update","signalTypeLabel":"Feature","sourceUrl":"https://www.harness.io/blog/the-contained-pen-test-proving-resilience-without-widening-the-blast-radius","publisher":"harness.io"}],"attribution":{"source":"Spyingbee","url":"https://spyingbee.com/updates/harness","citation":"Spyingbee, \"Harness updates\", https://spyingbee.com/updates/harness (retrieved 2026-10-08)"}}