Pillar 1 · Readiness X. Design Week 2026

AI readiness, workflow and knowledge systems

What it takes to move from using AI to building with it

What does it take to close the gap between using AI and building with it?

From using AI to building with it, in four layers

In this report
02The argument
03Change Tourism Austria
04Marketing Greece
05Stockholm Business Region
06Strategic recommendations
07Re-enter the zones
The argument

From using AI to building with it

Where adoption sits today

While the vast majority of destinations now incorporate AI into their daily routines, a significant disparity remains regarding how deeply they embed these tools into their core functions. Data from McKinsey indicates that 88% of organisations use AI on a day-to-day basis. In contrast, a mere 21% have fundamentally rebuilt their operational workflows around these new capabilities. This widespread adoption confirms that the technology is now firmly established across the sector. The remaining challenge involves the comprehensive redesign of daily work processes to fully absorb these advanced tools.

0%
use AI day to day
0%
have rebuilt a workflow around it

McKinsey, State of AI, 2025

The consequences of settling for surface-level adoption show up clearly in recent industry metrics. The GenAI Divide study from MIT NANDA reveals that 95% of enterprise artificial intelligence pilots produce minimal measurable returns. The underlying technological tools function highly effectively on their own merit. Organisations must therefore build deeper structural integrations to generate financial returns. Securing minor time savings on individual tasks provides immediate satisfaction while overall metrics for output and quality remain entirely unchanged. The most successful teams achieve significant value by redesigning their entire workflow to accommodate new capabilities. This comprehensive structural overhaul demands considerable time and deliberate effort to embed these systems deeply into daily team routines.

From individual to team AI

Individual use of AI naturally produces highly varied outputs. Team members apply their unique context to their daily requests to create results that diverge across the wider organisation. Destinations can bridge this gap by transitioning their technology from individual experimentation into a unified team resource. This fundamental shift requires teams to centralise their shared knowledge bases alongside fully documented workflows and an accessible brand voice. Advanced systems generate highly consistent deliverables when they receive unified inputs from these central repositories. Teams must therefore diligently document their procedures while continually maintaining their knowledge base and updating their brand guidelines to reflect ongoing evolution.

Workflows are becoming agentic

The upcoming operational shift revolves around the increasing levels of autonomy these systems carry within destination workflows. The process recognises three distinct stages of integration that clearly define how teams interact with the technology.

question_answer
Level 1
Answers
AI responds, human reviews and acts
approval
Level 2
Decisions
AI proposes, human approves
bolt
Level 3
Autonomy
AI acts within bounds, humans sets the parameters

Knowledge has become infrastructure

AI relies on clear contextual boundaries to maintain strict accuracy and relevance. Providing thorough context enables the technology to consistently generate reliable and strongly aligned content. Organisational knowledge currently functions as a vital operational infrastructure that actively powers these advanced digital tools. This essential infrastructure consists of four distinct components that guide the entire system. Brand knowledge encompasses the core strategic positioning of the destination. Approved sources represent the specific data sets and verified references the system actively utilises to build factual responses. Workflow rules clearly define how tasks progress collaboratively through the wider team. Editorial standards contain the detailed style guides and precise tone of voice requirements. These four elements combine seamlessly to ensure every generated output perfectly matches the high publication standards of the destination.

Judgement is the scarce skill

The technical barriers to adopting advanced digital tools have largely disappeared across the broader industry. Almost any team member can generate initial responses and construct effective operational workflows when the proper structural foundations exist. Human judgement now represents the most valuable and scarce resource within this modern working environment. Destination teams require this critical judgement to direct the technology toward highly productive outcomes and accurately evaluate the generated results while taking full ownership of the final product. Basic technical fluency rapidly becomes a standard commodity across the global workforce.

A structural shift, not a trend

Traditional trends typically rise to a peak before eventually returning to their original starting point. AI establishes a permanent foundation that fundamentally elevates all subsequent work. Every successful integration permanently raises the operational baseline for the entire organisation. Teams adopting shared knowledge bases maintain these vital resources indefinitely to support their daily tasks. Destinations deploying autonomous systems for routine administration naturally continue to rely on this automated efficiency. This newly established operational standard provides a lasting platform for continuous development.

Bolted on, or rebuilt

Switch between the two and watch what happens to the four layers.

AI sits on top · ceiling
AIStrategy
AIWorkflows
AITeam
AIData

AI layered on top. The four layers below stay as they were, which sets a ceiling.

Every destination marketing organisation must therefore actively weave these technologies deeply into their existing strategies, team structures and operational data. Adding new tools purely as a surface layer strictly limits the potential for long-term organisational growth. Teams achieve continuous compounding value by fundamentally rebuilding their core structural foundations around these advanced capabilities.

From adoption to readiness

The collaborative method established by the room progresses through three distinct stages of auditing alongside adopting and augmenting. The initial audit phase involves a comprehensive assessment to identify current team capabilities across existing tools and skills while locating operational gaps. Teams then progress to the adoption phase by integrating advanced technologies into everyday workflows supported by properly structured foundational knowledge. The final augmentation stage requires teams to fundamentally redesign their core work processes to maximise the new possibilities these systems create.

This process of augmentation actively reshapes the broader work environment to elevate human potential. Standard automation focuses entirely on executing standalone tasks to achieve basic efficiency. Augmentation directly empowers team members to handle increasingly strategic and highly valuable responsibilities. This operational shift successfully compresses the routine administrative layer to create vital space for complex problem-solving. Industry metrics clearly reflect this important transition across the global sector. Data from late 2025 confirms that the integration of augmentation strategies officially surpassed standard automation practices.

The augmentation tipping point
Late 2025
Automation45%
Augmentation52%

Augmentation has crossed past automation. The dashed line marks where automation stops and augmentation keeps going.

Case study

Change Tourism Austria

Building industry-wide AI capability through a community-first platform.

2,500
community members
35
AI use cases in the library
130
students at the latest hackathon

Austria Tourism · Change Tourism Austria platform

Austria Tourism's response to AI took the form of a community-first platform for the wider Austrian tourism industry. Change Tourism Austria started with a small group of early followers and grew over three years into a community of 2,500 members, with smaller groups inside the larger one driving the bulk of the knowledge sharing.

The platform sits on a three-pillar AI strategy. Guest experience, where AI is now in the conversation a visitor has before arriving and the destination has to be readable by the systems shaping that conversation. Workflows, where automation is positioned as relief for capable people working in a sector short on talent. Skills, where 68.8% of national tourism organisations cite enablement as the biggest barrier to AI, ahead of budget and data combined.

Why the response had to be community-first

The three shifts add up to a problem no single organisation can solve. A destination can publish its own data and automate its own processes. An industry only becomes AI-ready when thousands of operators, regions and small teams move at once. Two principles hold the community together. Access to the tools is the first step. Agency is the second, with people needing to feel safe enough to try AI in their work without fearing automation as a consequence.

What the platform offers

The platform runs on several formats, each reshaped around what the community is actually using.

podcasts
AI Radar

A monthly thirty-minute briefing on what is new in AI for tourism, free to join.

groups
Hackathons

Student teams paired with industry challenges. The recent InnoDays event produced 24 prototypes from 130 students across five themes.

library_books
AI Challenges library

35 use cases focused on messy reality, where what did not work matters as much as what did.

forum
Stammtisch mini communities

Smaller groups meeting regularly, where by the third meeting the polite questions have moved aside and the real conversations begin.

How the work is measured

Measurement runs on OKRs tied to community signals, with analytics on stories shared, topics that resonate and engagement that grows over time. The platform is treated as ongoing infrastructure, with each format reshaped around what the community is actually using.

Case study

Marketing Greece's AI playground

Building micro-tools that help the Greek tourism sector meet AI.

2 months
from start to shipped
Vibe Coding
low-cost tools, fast iteration
RAG
grounded in the team's content

Marketing Greece · a working platform of micro-tools

Marketing Greece operates as a private-sector fund and company at national level, with the Greek tourism industry as its stakeholders. The brief is to point the way, lead and bring the rest of the sector along. The team's response to AI took the form of a playground. A working platform of micro-tools designed to help small and medium tourism businesses meet AI without having to figure out the technology from scratch.

The working position is that this is a literacy challenge at heart. Greek tourism businesses are mostly small family operations without in-house AI knowledge and they need a way to meet the tools in a setting where they can build their own understanding. The platform is treated as enablement, with the goal of guiding businesses toward useful AI.

The tools inside the playground

The platform shipped in two months using vibe coding and low-cost tools, with several applications inside it.

shareSocial Lab

Generating image, text or quizzes for social posts.

reviewsReview Hub

Analysing review patterns and generating professional responses.

searchSEO Hub

Inspecting, briefing and writing optimised content.

securitySecurity Scanner

Identifying vulnerabilities for an AI-powered security audit.

imageBanner Studio

Generating on-brand campaign banners, hero images and social creatives.

photo_libraryContent Library

Browsing official tourism imagery from Marketing Greece's asset library.

Grounded by data

The fundamental principle running through every tool is the active grounding of information. A RAG sourced directly from the discovergreece.com platform anchors the AI firmly within the team's own approved content. A separate RAG handles all internal operational data to support daily administrative tasks. Team members treat every generated output as an initial draft requiring continuous human oversight across every piece of content.

The adoption story and the lesson for DMOs

The platform shipped before the marketing plan and uptake is still being worked through. Early feedback from hoteliers has been positive on the content creator. The next phase involves closer engagement with the businesses using the tools to understand which parts of their daily operations the platform should expand to support next.

AI literacy, treated as a competitive advantage the destination supports, is the framing other DMOs can take from this. Building micro-tools that let businesses meet AI in a low-stakes way moves the sector forward.

Case study

Stockholm Business Region's path to internal transformation

Moving from bottom-up experimentation to top-down strategic alignment.

Phase one
Bottom-up

Get colleagues curious. See what AI can do at an individual level.

arrow_forward
Phase two
Top-down

Get the management team on board so the 2027 priorities are set with AI woven through.

60% of staff already engaged with AI in some form, with 20 to 30% still standing at the platform

Stockholm Business Region operates as a fully owned municipal company combining tourism with inward investment. The DMO is responsible for attracting investments alongside talent and visitors across the middle of Sweden. The team has spent the past year actively determining how AI serves an organisation managing this combined remit.

The operational journey began about a year ago by actively encouraging a bottom-up approach to technological innovation. Leadership invited colleagues to explore their curiosity and discover the capabilities of these new tools at a purely individual level. This initial instinct proved highly effective across the wider organisation. Internal research showed that around 60% of the staff already engage with the technology in some capacity while another 20% to 30% are actively preparing to join them. The strategic landscape at the leadership level presented a uniquely different scenario. The management team found themselves actively moving toward the starting point of this major transformation. Leadership fully recognised the necessity of this transition while continuously working to build a complete understanding of how these systems will fundamentally change the entire organisation.

The shift to top-down

For a stretch, the experimental phase was pushed hard and AI was being put everywhere across the organisation, which built visible momentum and capability. With the experimental phase well underway, the team has now slowed down to focus on the missing piece. Getting the management team on board so the strategic priorities for 2027 are set with AI woven through them from the start.

From divergent output to aligned output

The organisation is moving from individual AI use, where each person's output diverges, to aligned organisational use, where the same context produces the same kind of work across the team. Getting there requires the system level. Bringing people into a shared way of working, with the management team backing the structure, is the work that turns experimentation into integration.

The lesson for DMOs

This path presents a highly authentic example of how technological integration successfully unfolds across an organisation. Initial bottom-up experimentation actively builds team curiosity alongside highly visible early victories. Strategic top-down alignment subsequently transforms those individual wins into meaningful structural change. Organisations absolutely require both distinct phases to achieve their desired operational outcomes. Dedicating comprehensive time to each stage directly maximises the final value the team ultimately achieves.

Strategic recommendations

What the room asked, and where it landed

The questions DMOs were working through, with the answers the conversation settled on. Six recommendations follow.

01

How fast should we move with AI?

The honest answer is faster than feels comfortable and slower than feels reckless. Moving too fast leads to AI workflows producing off-brand or inaccurate work, which damages trust. Moving too slowly leaves the destination behind teams who are integrating AI into how the work happens, which compounds against you. The framing the room landed on was to think in terms of readiness. Move at the pace your foundations support, and build those foundations as quickly as the team can absorb the change. Audit where you are. Adopt the tools that fit. Augment the work itself once both are in place.

02

What does AI readiness actually mean?

expand_more

AI readiness is the state where a team has the foundation to absorb AI into how the work runs, with consistent output and shared knowledge holding the integration together. Four conditions sit underneath it. An inventory of where AI is already being used, including the informal use that has not been documented. A shared knowledge base the AI can read, so outputs stay aligned across the team. At least one workflow redesigned with AI woven in from the start. Someone owning the AI work as part of their formal role. When the four hold together, the team can scale AI use without producing inconsistency. Until then, the work is still adoption.

03

Where do we start when there is no obvious owner?

expand_more

The most successful teams establish clear ownership early in the process even when the initial role remains relatively small in scope. Appointing a part-time coordinator with the explicit brief to inventory current usage alongside building the shared knowledge base and identifying the first workflow to redesign provides an excellent starting point. This dedicated role must report directly to senior leadership to ensure the initiative receives the necessary backing to drive ongoing progress. Centralising this vital responsibility ensures the broader team remains entirely unified in their daily approach to successfully capture the valuable integration gains the new structure actively provides.

04

How do we deal with public sector legal and procurement constraints?

expand_more

The teams who succeed treat legal and procurement as partners from the start. Bring them into the conversation when you are scoping the AI work, not when you are trying to deploy it. Give them time to develop frameworks for AI tools the same way they developed frameworks for cloud services or third-party platforms. Many of the constraints are about data handling and supplier accountability, which are solvable problems when the legal team is involved early. The constraints become blockers when the legal team is presented with a finished deployment plan and asked to sign off in a week.

05

What is the difference between automation and augmentation, and which matters more?

expand_more

Automation eliminates routine tasks while augmentation actively elevates human potential. Both approaches provide distinct utility while augmentation delivers the most substantial long-term value. Automating existing processes captures a singular efficiency saving for the team. Augmenting team members fundamentally expands their capabilities across every assignment to generate continuously compounding returns. Industry consensus during the sessions indicated that this strategic balance has already shifted. Data from late 2025 confirms that the adoption of augmentation officially surpassed standard automation practices. Teams concentrating on automation optimise their current operational workflows. Meanwhile, organisations prioritising augmentation proactively reshape the entire scope of their future capabilities.

Recommendations to take into the second half of the year
01

Audit current use before adopting anything new

Half the gains from AI come from organising what people are already doing. The unmapped uses inside the team are the most common source of inconsistency and brand drift. An honest audit, conducted without judgement, is the first investment.

02

Build a shared knowledge base before scaling AI use

Brand voice, approved sources, workflow rules and tone of voice need to live somewhere the AI can read them. The same model produces aligned output when the inputs are aligned. The work is organisational.

03

Pick one workflow to redesign, not several

Redesigning several workflows at once dilutes the work and produces half-finished outputs. Pick one, redesign it properly, document what worked, then move to the next.

04

Assign ownership, even part-time

The most reliable predictor of successful AI integration involves assigning clear ownership of the transition to a dedicated individual. This essential role can begin as a part-time responsibility provided it carries some level of authority to execute decisions across all departmental teams.

05

Treat legal and procurement as partners, not gates

Destination leaders must bring key stakeholders into the process at the earliest opportunity to provide them with sufficient time for meaningful contribution. Teams should actively frame this collaborative work as a partnership while treating all operational constraints as highly productive design parameters.

06

Plan for augmentation alongside automation

The compounding value sits in changing what people can do. Removing tasks captures the saving once. Augmenting the person keeps paying back across every task they touch. Design the AI integration around what the team becomes able to do over time.

lock

Log in to continue reading.

From the room lockMembers

Re-enter the zones from XDW 2026

Each zone took the readiness question through a different posture. The summaries below capture what happened in the room and where each conversation landed. The portal links open the immersive zone experience built during the event.

science

The Lab

What happened

The Lab took the readiness question into a hands-on session on building knowledge systems. The premise was that any workflow built around AI starts with knowing where to find the information the AI needs, which means building the underlying knowledge system first. The session walked through setting this up in practice, using project knowledge in Claude, with MCP connections linking it into wider workflows.

The deeper layer came in the discussion of skills. A skill is a saved package of knowledge, written as a plain markdown file, that holds the steps and the context for a specific workflow. When activated, the AI applies that knowledge to whatever task it is given. The session walked through how an advisory time-logging workflow runs in practice, with the agent reading the request, applying the skill and producing a structured outcome with human sign-off at the right points.

Explore The Labnorth_east
hub

The Strategy Room

What happened

The Strategy Room explored methods to progress alongside an evolving strategy. Taking immediate action delivers greater value than waiting for complete certainty. Three distinct operational patterns emerged from these insights. Teams can create visible wins by solving specific workflow bottlenecks with new technology to drive internal adoption. Providing personal attention to small groups effectively builds confidence among hesitant colleagues. Senior leadership must actively support these initiatives to ensure widespread engagement across the entire workforce.

The room also worked through the constraints public organisations face. Procurement processes that block obvious tools. Approval cycles that take months for things teams have already been doing informally. The destinations getting traction are working with their legal and procurement teams as collaborators brought in early enough to develop frameworks together.

Explore The Strategy Roomnorth_east
forum

The Debating Room

What happened

The Debating Room ran two related topics consecutively to explore current operational challenges. The first discussion examined whether the primary barrier to integration involves human factors or technological limitations. Participants concluded firmly that the main challenges revolve entirely around people. The underlying technology remains highly accessible to the entire team. Successful adoption relies heavily on building confidence alongside providing a space for practical experimentation. A highly useful thread subsequently emerged regarding brand integrity and information quality. Advanced systems require a definitive version of the brand guidelines to successfully generate aligned content. The second topic addressed the relationship between increased content production and overall quality.

The room actively challenged the initial premise by acknowledging that varying content quality has always been a constant factor across the industry. The participants instead focused their attention on the critical balance between output volume and professional judgement. Advanced systems can easily generate eighty pieces of content in the exact timeframe a human writer traditionally requires to produce a single piece. Teams must therefore actively build the necessary editorial capacity to carefully review all generated material.

Explore The Debating Roomnorth_east
support_agent

The Advisory Clinic

What happened

The Advisory Clinic explored the transition from individual technology use to unified team capability. Knowledge organisation emerged as the dominant operational pattern. Teams of individuals applying their own context struggle to scale as outputs diverge and institutional knowledge remains isolated. Successful destinations actively build shared knowledge programmes that centralise information across the entire team. These central resources consolidate brand guidelines alongside editorial standards and factual references into one highly structured format.

A second pattern revealed that current operational challenges remain primarily foundational. Most destinations possess access to similar digital tools while comprehensive preparatory work clearly distinguishes the highly successful teams. Specific platform choices play a minor role because a robust knowledge base effectively supports whichever system the DMO ultimately selects.

Explore The Advisory Clinicnorth_east
arrow_backBack to overview Continue to Governancearrow_forward
lock

Log in to continue reading.

chevron_right
01 · Overview
keyboard_arrow_leftArrow keys · swipe · tap a dotkeyboard_arrow_right
01 / 07