AI is moving beyond the experimentation phase.
Across industries, organizations are making statement investments in AI to boost productivity, accelerate decision-making and uncover new opportunities for growth. Yet many leaders are still searching for a clear path that takes them from pilot projects to measurable business outcomes.
While nine in 10 (88%) organizations are testing AI, only 19% report any meaningful business gains, according to McKinsey’s The State of Organizations 2026 report. The challenge here remains deploying and scaling AI in a way that’s trusted, governed, and delivers tangible value.
That roadblock shaped many of the conversations at the latest Alteryx on Tour event, in London, UK, where customers, partners and industry leaders shared their perspectives on what it actually takes to move AI from exciting idea generator to production-ready technology. Three themes stood out.
1. The defining challenge: Trust
The opportunity AI presents is enormous – and so too are the risks of deploying it without the right governance and controls.
As AI becomes embedded into more business-critical processes, trust has rightly become the defining challenge. Leaders need confidence that AI-generated outputs can be understood, validated and governed before they’re used to support important decisions, and at scale.
During his keynote, Alteryx CEO Andy MacMillan shared a simple scenario that captured AI’s biggest stumbling block – its probabilistic nature. After receiving incorrect financial advice from a large language model, he asked another LLM why the answer was wrong.
Its response? “Well, LLMs are often wrong.”
If AI can’t be trusted, then it can’t be operationalized.
To address this, he introduced the VURA framework: AI must be Visible, Understandable, Repeatable and Auditable to give organizations confidence that every insight is transparent, explainable and grounded in trusted business logic.
That’s especially true as organizations begin adopting agentic AI. Unlike traditional chatbots that answer basic questions, AI agents are expected to analyse information, act, and make decisions autonomously. Without trust, governance and transparency, that potential quickly turns into risk.
The concern is widespread. Only 3% of analysts are comfortable with fully autonomous AI, highlighting just how important trust, transparency and governance have become as organizations race to scale AI.
Of course, that doesn’t mean organizations should stop using AI or large language models. The opportunity isn’t AI replacing the trusted workflows that are already powering insight-rich dashboards, mission-critical reports, and enterprise-grade automations. The opportunity is extending these workflows into AI.
Existing analytics workflows and business logic become the foundation for AI, transforming workflows from static assets into interactive ones. Business users can engage with them using natural language, while AI reasons over business logic rather than generating answers in isolation. The result is faster decision-making without sacrificing transparency or traceability.
2. In the LLM age, business expertise matters
Organizations have spent years investing in modern cloud data platforms, analytics workflows and governance models. The next step would not be to swap out those investments for AI, but connect AI to the trusted data, business logic and expertise that already drive decisions.
That matters because AI can’t automatically infer an organization’s financial controls, operational policies or regulatory requirements. To scale successfully, AI needs business expertise embedded into the way it is built, governed and used.
This message resonated throughout Alteryx on Tour.
During the UBS session, the bank highlighted the importance of “governing the system, not the user”. The aim is not to restrict business teams from using AI and analytics, but to give them a secure, governed foundation where they can contribute safely and effectively.
That means enabling a range of users – from citizen builders to everyday business users and advanced practitioners – to apply analytics where they understand the context, while IT maintains the controls, standards and guardrails needed to scale with confidence.
Organizations that successfully scale AI aren’t choosing between business and IT – they’re bringing both together to combine trusted data, business expertise and governance into better outcomes.
Kickstart your journey: Download our AI-ready Starter Kits today
3. Solve for business (not tech) problems
As AI continues to dominate headlines, many of the most compelling stories focus on solving practical challenges.
One standout example came from the Marine Conservation Society. With a smaller data and analytics team, the charity built an AI-powered model to measure the impact of public policy on beach litter, helping drive evidence-based environmental advocacy.
The team is using analytics and AI to transform complex findings into plain-language reports before making insights available through a conversational AI agent. This allows stakeholders to ask follow-up questions in natural language – extending access to insights without placing additional demands on the analytics team.
That same principle was reflected across the event. Rather than pursuing automation and AI for its own sake, organizations are embedding the technology into workflows that drive their businesses.
Card retailer UK Greetings has fully automated its business intelligence reporting and is now scaling analytics use cases across its supply chain, from reducing waste to optimizing sales opportunities.
Marex shared how it’s using Alteryx One to drive continuous self-automation, while Zenith showcased how automation is transforming its finance operations at scale.
Organizations aren’t starting with AI and searching for a use case. They’re starting with a business challenge and using analytics, automation and AI to solve it faster and smarter.
What’s next?
The conversation around AI and agents is evolving rapidly.
Organizations are moving beyond experimentation and asking a more important question: How do we scale responsibly?
The answer is building trusted foundations, embedding business expertise, and empowering the people closest to the work to innovate with confidence.
As AI continues to move at pace, the organizations creating the greatest value won’t necessarily be those adopting it the fastest. They’ll be the ones that build trust first – combining data, business expertise and AI to deliver meaningful business outcomes.