Key Takeaways
- Scaling AI requires organizational change, not just new technology.
- Most enterprises are still moving from experimentation to operational transformation because governance, workflows, and decision rights are harder to redesign than deploying AI tools.
- AI creates uneven productivity gains, making it essential to identify and remove new organizational bottlenecks.
- Human oversight, governance, and cross-functional alignment remain critical as AI becomes part of everyday work.
- Organizations that redesign how teams operate will capture more value than those focused solely on individual productivity.
Most organizations have already answered the first AI question: can it make people more productive?
They know artificial intelligence can accelerate content creation, summarize meetings, analyze data, and automate repetitive work. The harder question is how to turn those isolated productivity gains into a lasting competitive advantage across the business.
To understand what that transition looks like in practice, Activate Summit 2026 brought together technology and marketing leaders from Iterable, The Washington Post, and Dropbox. Samya, CTO at Iterable; Vineet Khosla, CTO at The Washington Post; and Joyce, Head of Marketing at Dropbox, shared how their organizations are approaching the next stage of AI adoption, from redesigning workflows and decision rights to introducing new roles and governance.
The discussion offers a practical framework for leaders who have already experimented with AI and are now asking a different question: what does an AI-native organization actually look like?
Editor’s note: Watch the full Activate Summit 2026 session, “The AI-Native Organization: Structure That Scales Intelligence,” on demand. Link to come.
AI Doesn’t Fail Because of Technology
Much of the conversation around AI focuses on speed and productivity. Teams can generate more content, write more code, and complete routine work in a fraction of the time.
Joyce argued that increasing individual productivity is only the beginning. As she put it, “AI is only going to amplify your gaps.”
Scaling AI therefore requires organizations to address the weaknesses that productivity tools alone cannot solve, including governance, accountability, decision rights, and the relationship between human judgment and AI systems. The challenge becomes clear as soon as work crosses team boundaries.
AI can generate a campaign plan in minutes. It cannot align five stakeholders on priorities, resolve competing objectives, or build consensus across functions. Those responsibilities remain firmly in the hands of people, making organizational design the limiting factor rather than the technology itself.
That distinction explains why many AI initiatives plateau after early success. Teams adopt new tools, but existing approval processes, governance models, and operating structures remain unchanged. Productivity improves within individual functions while the organization continues moving at the same pace.
For leaders, that changes where transformation begins.
Instead of asking which AI tool to deploy next, organizations should ask:
- Which workflows create the most friction?
- Where do decision rights move between people and AI?
- Which approvals still require human judgment?
- What organizational changes are preventing AI from creating broader business value?
Takeaway: AI scales through organizational redesign. Technology creates new capabilities, but governance, workflows, and decision-making determine whether those capabilities produce lasting business impact.
Most Organizations Are Still in the Middle of the AI Adoption Curve
Despite rapid advances in AI, very few organizations have reached enterprise-wide transformation.
Joyce described most established companies as sitting in the middle of the AI adoption curve. The obstacle is not a lack of ambition. It is the operational complexity involved in changing how an entire organization works. Moving from successful pilots to measurable business outcomes requires teams to align around shared definitions of success, documented workflows, and clearly defined responsibilities for both humans and AI.
The panel highlighted several questions leaders should answer before expanding AI further:
- Does the organization share a common definition of success?
- Have critical workflows been mapped and documented?
- Are decision rights clearly defined for people and AI?
- Do governance processes support faster execution instead of slowing it down?
Vineet offered a practical example from The Washington Post. While the organization as a whole sits between adoption and optimization, different functions have progressed at very different speeds. Finance has transformed invoice reviews through AI, while comment moderation has fundamentally changed how journalists engage with readers by creating healthier conversations and encouraging more direct interaction.
Those differences illustrate an important reality. AI adoption does not happen evenly across an organization. Some workflows create immediate value, while others require more experimentation before meaningful transformation becomes possible.
Joyce summarized the broader lesson simply: AI amplifies existing organizational strengths and exposes existing weaknesses. Gaps in governance, unclear ownership, and disconnected workflows become more visible as AI scales.
Takeaway: Most organizations are not struggling with AI capability. They are working through the operational changes required to move from isolated pilots to organization-wide transformation.
AI Creates New Bottlenecks, Not Just New Productivity
Early AI adoption focused on individual productivity. Content teams could generate more campaigns, engineers could write code faster, and analysts could automate reporting.
The next challenge is that those gains rarely arrive evenly across an organization.
Joyce compared it to software development. A 10x engineer does not automatically create a 10x product if product management, design, or other functions continue operating at the same pace. Marketing experiences the same pattern. Content production accelerates, but brand review, copy quality, tone, and cross-functional approvals become the new constraints.
That shifts the role of leadership.
Instead of celebrating isolated productivity gains, leaders need to identify where work is now accumulating and redesign those parts of the process.
Common bottlenecks include:
- Brand review and approval.
- Cross-functional alignment.
- Governance and compliance.
- Human quality control.
- Decision-making across multiple stakeholders.
Removing those constraints creates more value than simply asking every team to generate more work. AI changes where the organization slows down, and those new bottlenecks become the next opportunity for improvement.
Takeaway: AI does not improve every workflow equally. Organizations create the greatest value by identifying where work now stalls and redesigning those constraints.
Build Around Organizational Principles Before AI Projects
The Washington Post approached AI adoption differently than many organizations. Before building products or deploying new tools, the leadership team agreed on the principles that would guide every AI initiative.
Vineet Khosla explained that the company aligned around three priorities:
- AI must improve the customer experience.
- AI must preserve editorial integrity.
- AI must generate business value.
That alignment simplified later decisions because every proposed project could be evaluated against the same objectives. Teams no longer debated whether AI should be used. They focused on whether a specific use case supported those principles.
The process also created broader organizational buy-in. Once leadership agreed on the goals, product, engineering, newsroom, and business teams could move forward with a shared understanding of what success looked like.
For organizations beginning larger AI programs, establishing common principles may be one of the highest-leverage decisions they can make. It creates consistency across projects while giving individual teams flexibility in how they solve problems.
Takeaway: Shared principles create better AI governance than isolated project decisions. Alignment at the leadership level helps organizations move faster while maintaining consistent standards.
Design Teams Around Adoption, Not Just Delivery
Governance alone does not create transformation. Teams also need an operating model that helps AI move from experimentation into everyday work.
The Washington Post built a dedicated AI pod responsible for developing shared capabilities while partnering closely with business functions across the organization. Rather than operating as an isolated innovation team, the AI specialists worked directly with product, newsroom, finance, customer service, and other groups to solve real operational problems.
The structure created a continuous feedback loop. Partner teams brought business requirements and operational context, while the AI pod developed solutions and refined them based on real-world use. Different functions were also allowed to adopt AI at different speeds instead of forcing a single process across the entire organization.
Comment moderation illustrates how that model evolved. The Washington Post initially kept humans involved in reviewing AI decisions while the team built confidence in the system. As accuracy improved and trust increased, moderation became increasingly autonomous, allowing journalists to spend less time managing hostile discussions and more time engaging with readers.
The same philosophy extended beyond customer-facing products. The company deliberately applied the same AI platform across finance, HR, legal, customer service, and editorial workflows so every function could benefit from the organization’s investment rather than limiting AI to a single flagship product.
Takeaway: AI scales through operating models that connect technical expertise with business teams. Dedicated AI capability, shared governance, and continuous feedback help organizations move beyond pilots into everyday operations.
The Go-to-Market Engineer Is Becoming a Core Marketing Role
AI has lowered the barrier to building automations, agents, and workflows. That does not mean every marketer suddenly becomes a software engineer.
Joyce argued that marketing organizations increasingly need a technical profile embedded within the team that can bridge strategy and execution. She calls this role the go-to-market engineer. Unlike traditional marketing operations, this person designs AI agents, connects systems, builds workflow automations, manages integrations, and helps ensure AI outputs meet quality standards.
The role is emerging quickly. Joyce noted that LinkedIn job postings for go-to-market engineers have grown from almost none two years ago to thousands today, reflecting how organizations are restructuring around AI rather than simply purchasing new tools.
Samya reinforced the idea by sharing that Iterable’s marketing organization already includes a GTM engineer. The impact is measured less by the technology itself than by how much faster the marketing team can move because technical expertise sits alongside marketers instead of outside the department.
The role represents a broader shift in marketing.
- Build workflows instead of completing repetitive tasks.
- Connect systems instead of managing disconnected tools.
- Create AI agents that support teams across the organization.
- Improve AI quality through continuous testing and refinement.
Takeaway: AI-native marketing teams increasingly combine strategic marketers with embedded technical builders who can translate ideas into scalable workflows.
Hire for Systems Thinking, Not AI Expertise
AI skills are changing quickly. Foundational thinking changes much more slowly.
Rather than hiring people because they have “AI” in their job title, the panel encouraged leaders to look for systems thinkers who understand how work flows across an organization and can orchestrate people, technology, and AI together.
Joyce identified two capabilities becoming more valuable:
- Systems thinking.
- Orchestration.
As repetitive execution becomes increasingly automated, marketers spend more time coordinating workflows, managing AI agents, and improving processes across teams. Individual contributors become managers of AI agents in much the same way people managers coordinate employees.
Joyce offered a useful comparison.
Onboarding an AI agent follows many of the same principles as onboarding a new employee. Teams provide context, define goals, establish constraints, review outputs, and offer iterative feedback until performance improves.
Vineet expanded on that idea by emphasizing the continued importance of expertise. AI can amplify the work of engineers, journalists, accountants, and marketers, but only if those experts understand what good work looks like in the first place.
As he explained:
“You need experts a whole lot more.”
Takeaway: AI increases the value of domain expertise. Organizations should hire people who understand systems, exercise sound judgment, and can direct AI toward meaningful business outcomes.
Content Strategy Is Changing Alongside Search
AI is also changing how customers discover information.
Joyce noted that more than half of searches now begin with an LLM, creating a discovery experience that blends earned media, owned content, product pages, customer communities, and other sources into a single response.
That shift changes how marketing content should be written. Content teams now need to think about two audiences:
- People reading the content.
- AI systems retrieving and synthesizing it.
That requires clearer structure and stronger specificity. Instead of relying on broad statements, content should:
- Break complex ideas into clearly defined sections.
- Support claims with specific proof points.
- Use precise language instead of vague marketing terminology.
- Organize information so AI systems can retrieve it accurately.
The change also affects organizational structure. Brand marketing and performance marketing can no longer operate independently if AI systems are pulling information from every customer touchpoint simultaneously. Consistent messaging across every channel becomes part of search strategy, not simply brand governance.
Takeaway: AI search rewards clear, well-structured content supported by evidence. Consistent messaging across teams becomes increasingly important as LLMs synthesize information from multiple sources.
AI Changes the Operating Model, Not the Mission
The panel closed with an important distinction.
Marketing’s core responsibilities remain the same. Organizations still need to build customer trust, generate revenue, and strengthen their brand. What changes is how those outcomes are achieved and how leaders measure progress.
Joyce argued that customer journeys have fundamentally changed as AI reshapes discovery and consideration. Traditional traffic and conversion benchmarks no longer capture every step because customers increasingly begin their research inside AI systems rather than conventional search engines.
Vineet offered a useful counterbalance. The Washington Post still measures success using many familiar business metrics because, in his words, “business is the business.” Even organizations leading AI adoption are still learning how measurement should evolve.
That tension reflects the broader state of AI adoption today. The technology is advancing rapidly, but the operating playbook is still being written.
The discussion points to four priorities for leaders:
- Redesign workflows before buying more AI tools.
- Remove organizational bottlenecks instead of chasing isolated productivity gains.
- Build teams that combine technical capability with business expertise.
- Experiment continuously while documenting what works.
Organizations that embrace that mindset will be better positioned to adapt as AI continues to reshape how work gets done.
Editor’s note: Watch the full Activate Summit 2026 session, “The AI-Native Organization: Structure That Scales Intelligence,” on demand. Link to come.
Frequently Asked Questions (FAQs)
What is an AI-native organization?
An AI-native organization integrates AI into everyday workflows, governance, and decision-making rather than limiting it to isolated experiments. The panel emphasized that organizational structure, leadership alignment, and clearly defined decision rights are as important as the technology itself.
Why do AI projects struggle to scale?
According to the panel, AI projects often stall because organizations fail to redesign workflows, governance, and decision-making. AI amplifies existing organizational gaps instead of solving them automatically.
What is a go-to-market engineer?
A go-to-market engineer is a technical role embedded within a marketing organization that builds AI agents, workflow automations, integrations, and supporting infrastructure. The role helps marketing teams adopt AI more quickly while maintaining quality and governance.
How should companies hire for AI?
The speakers recommended hiring for systems thinking, domain expertise, and the ability to orchestrate people and AI together rather than focusing exclusively on AI-specific job titles or technical skills.
How is AI changing content strategy?
As more customer discovery happens through LLMs, content needs clearer structure, stronger semantics, supporting evidence, and consistent messaging across channels. Organizations should create content that serves both human readers and AI retrieval systems.