The era of AI washing is over

The period of rewarding businesses for simply announcing an AI initiative is drawing to a close, as noted in the latest AI Daily Brief. The increasing power and accessibility of open-source models, coupled with a more sophisticated discourse around cost, customisation, and routing, is demanding real substance over marketing spin. For business operators, this signals a critical shift: move beyond shallow AI implementations and focus on genuine workflow redesign to unlock real, measurable value.

The practice of AI washing — making exaggerated or unsubstantiated claims about the use of artificial intelligence to attract investment and customers — is becoming increasingly difficult to sustain. According to The AI Daily Brief, the market is maturing rapidly, moving away from rewarding hype and towards demanding demonstrable results.

Several factors are driving this shift:

• Maturing Technology: The proliferation of powerful open-source AI models means advanced capabilities are no longer the exclusive domain of a few large tech companies. This accessibility makes it easier to verify claims and harder to feign sophistication.
• Sophisticated Discourse: The conversation in boardrooms and technical teams has evolved. Simple chatbot integrations are table stakes. The real discussion now centres on more complex, value-driven concepts like:
  • Model Routing: Intelligently directing tasks to the most appropriate AI model based on complexity and cost.
  • Customisation: Using techniques like Retrieval-Augmented Generation (RAG) or fine-tuning to make models genuinely useful for specific business contexts.
  • Cost Management: Actively monitoring and optimising the cost of AI model usage (inference costs) to ensure a positive return on investment.
  • Organisational Redesign: Recognising that AI's true potential is unlocked by changing how work is done, not just layering new tools on top of old processes.

This evolution marks a clear transition from a phase of exploration and hype to one of practical, cost-conscious implementation.

The two eras of corporate AI

For business owners and operators, this shift from AI washing to practical application has profound implications. Your competitors are no longer just talking about AI; they are quietly integrating it into their core operations to create significant advantages. Ignoring this transition is a direct risk to your own competitiveness.

From technology problem to organisation problem

Previously, the main challenge was accessing AI technology. Now, with powerful models widely available, the challenge is organisational. Simply providing your team with access to a tool like ChatGPT Enterprise or Claude is not an AI strategy. It's a software purchase.

True value comes from fundamentally rethinking and redesigning your business processes around the capabilities of AI agents. This means moving from human-led tasks augmented by AI to AI-led workflows supervised by humans. This is a much deeper change that impacts job roles, team structures, and management practices.

Cost and complexity are the new frontiers

Treating all AI tasks the same is a recipe for wasted expenditure. Using a state-of-the-art model like GPT-4o to summarise simple meeting notes is like using a sledgehammer to crack a nut — expensive and inefficient. The mature approach involves building agentic systems that can make economic decisions, such as routing a simple data extraction task to a cheaper, faster model, while reserving the more powerful models for complex reasoning or creative tasks. Businesses that master this cost-performance balancing act will have a structural cost advantage over those who do not.

Productivity is not an automatic outcome

Productivity gains do not magically appear when you subscribe to an AI service. They are the result of deliberate workflow redesign. For example, instead of an accounts payable clerk manually checking invoices and entering data, a well-designed AI agent could:

1. Ingest an invoice from an email inbox.
2. Extract the relevant data (vendor, amount, due date).
3. Cross-reference it with a purchase order in your accounting system.
4. Flag any discrepancies for human review.
5. Stage the payment for approval.

This is not a simple chatbot query; it is a multi-step, automated process that fundamentally changes the nature of the work. This is where real productivity is unlocked, and it requires more than just a press release.

To move beyond AI washing and build a real competitive advantage, businesses should take a structured, practical approach.

1. Conduct a process-centric AI audit. Don't just list the AI tools you're using. Instead, map your core business processes (e.g., sales qualification, customer support, financial reporting) and identify the specific steps that are most repetitive, time-consuming, or error-prone. This is your target list for AI intervention.

2. Focus on workflows, not just tools. Frame the problem around the process. Instead of asking, "What can we do with Claude 3.5 Sonnet?", ask, "How can we reduce our client onboarding time from 3 days to 3 hours?" This focuses the effort on a measurable business outcome, where an AI agent might be part of the solution.

3. Develop a cost-aware implementation plan. Work with your technical team or an external partner to design systems that use a mix of models. Establish a 'router' or 'orchestrator' that selects the most cost-effective model for each specific task within a workflow. Start tracking 'cost per task' as a key performance indicator.

4. Invest in operational skills, not just technology. The biggest gains come from upskilling your team to work with AI agents. This includes training them on how to define tasks, structure data for the AI, supervise automated workflows, and handle exceptions. This is a shift from 'doing the work' to 'designing and managing the work system'.

5. Pilot, measure, and scale. Select one high-value, well-defined process from your audit. Build a pilot agentic workflow to automate it. Measure everything: time saved, error rate reduction, cost per transaction, and employee satisfaction. Use this hard data to build the business case for scaling the solution to other parts of the organisation.