The News
At KubeCon North America 2026, Manik Surtani, CTO and co-founder of the Agentic AI Foundation, spoke with ECI Research about the organization’s trajectory eleven months after its founding. Surtani outlined the foundation’s dual mandate: drawing high-momentum open source projects under its umbrella and building a global conference series (AgentCon) that collocates maintainer meetups with practitioner programming. The foundation is planning ten events for 2027, including flagship gatherings in San Francisco and London, with additional events in China, Japan, Australia, and India.
Analyst Take
The Foundation Model as Neutral Ground
The Agentic AI Foundation’s timing is not accidental. The agentic AI ecosystem is fragmented by competing frameworks, overlapping standards efforts, and vendor-driven narratives that each pull in different directions. A neutral foundation with real member funding and a project incubation mandate fills a structural gap that commercial vendors cannot credibly fill on their own. Surtani’s framing was explicit: this is not a pay-to-play vehicle where membership buys influence. That’s the right message, and it’s the message that will determine whether the foundation earns legitimacy or becomes another logo farm.
For ITDMs, the practical value is clarity. When the foundation eventually publishes reference architectures or certification frameworks, those outputs will carry more weight than any single vendor’s white paper precisely because they emerge from a multi-stakeholder process. The question is how long that process takes to produce anything actionable.
The Skill Gap Is the Real Bottleneck
Surtani flagged the talent shortage as critical, and the conversation around AI job titles echoed a pattern the industry has seen before: SRE, DevOps engineer, and now “FDE” (AI-focused developer engineer) are terms that get attached to wildly varying skill sets. Employers can’t reliably assess competence, and service providers are building proprietary certification programs in the absence of any standard. The foundation is being asked to help design some of those programs, which is a meaningful opportunity to shape what “AI-ready developer” actually means at an industry level.
ECI Research’s Google GovTech Survey found that 47.2% of respondents selected “Developer velocity and ease of integration” as the greatest weight in their final technical selection process, assuming baseline security and compliance requirements are met. That finding reframes the talent question: organizations aren’t just asking whether someone knows agentic AI in the abstract, they’re asking whether that person can make systems move faster and integrate cleanly. A certification framework anchored to velocity outcomes rather than theoretical knowledge would be far more useful than one that tests framework familiarity alone.
Sandboxes Are Necessary but Not Sufficient
The KubeCon keynote moment Surtani referenced, the industry conversation around sandboxing for agentic systems, captures a genuine technical inflection point. Sandboxes protect the perimeter. They do not address what happens when agents operate inside the perimeter with legitimate access to sensitive systems, data, and APIs. That’s the harder problem, and it’s largely unsolved. ECI Research’s Google GovTech Survey found that 31.8% of respondents cited “FedRAMP/compliance approval friction for AI vendors” as the single largest blocker preventing widespread AI adoption in developer workflows, a finding that points to the same core tension: security frameworks built for static software deployments are poorly suited to the dynamic, multi-agent architectures that are becoming standard.
For developers, this matters architecturally. Designing agentic systems today means making bets on security and isolation patterns that don’t yet have industry consensus. The foundation’s role in standardizing those patterns, if it can execute, would give developers something genuinely useful: a reference model for building agents that can clear compliance gates without bespoke security reviews for every deployment.
Looking Ahead
The Agentic AI Foundation’s first year has been about establishing credibility and pulling projects in. The next twelve months will test whether it can produce outputs that have operational weight: certification standards that employers trust, security reference architectures that compliance teams can point to, and project governance models that give enterprise adopters confidence in long-term support. The foundations that succeed in the open source ecosystem are the ones that solve real coordination problems rather than simply convening conversations. The Agentic AI Foundation has the funding and the membership base to do the former; the question is whether its leadership will resist the temptation to let the latter crowd out the former.
The competitive dynamic to watch is how hyperscalers and major platform vendors engage with the foundation over the next two to three years. Right now, member companies are funding the foundation because the ecosystem is too early for any single vendor to lock in. As agentic AI matures and commercial patterns become clearer, the incentive structure will shift. The foundation’s ability to maintain genuine neutrality while navigating increasing commercial pressure will define its long-term relevance, and its long-term relevance will define whether the industry gets shared standards or a fragmented collection of proprietary frameworks dressed up in open source language.
Stay Ahead of Application Development Trends
Get weekly analyst insights, research notes, event coverage, and AppDevANGLE updates delivered directly to your inbox.
Subscribe for Weekly Insights
Join technology leaders, practitioners, and GTM teams following the trends shaping modern software delivery.
Looking for deeper research access?
Explore ECI Research reports, survey insights, and market analysis through the ECI Research Portal.
