A new study from Descartes Systems Group shows a clear gap between AI interest and large-scale AI adoption in transportation.
 

The company's 10th Annual Global Transportation Management Benchmark Survey, released on September 16, 2026, found that transportation technology investment has increased significantly over the past decade. However, only 19% of shippers and 15% of logistics service providers say they are using AI at scale. Most organisations are still experimenting, piloting or deploying AI only in selected areas.

📈 Technology investment is increasing

The research shows that transportation technology has moved from being an operational tool to a strategic priority.
 

Among surveyed North American organisations, the share planning to increase transportation-management technology investment has risen from 53% in 2017 to 78% in 2026. Meanwhile, 85% now view transportation as either a competitive weapon or customer-service differentiator, compared with 65% a decade ago.

So the issue isn't a lack of interest in technology.

The challenge is turning that investment into connected, scalable operations.

🤖 AI adoption is still stuck between pilots and scale

The survey found that only 19% of shippers and 15% of LSPs have scaled AI across their operations.

For many companies, AI remains focused on individual use cases rather than being integrated throughout the transportation workflow.

This creates an important distinction:
 

Having an AI tool isn't the same as having an AI-enabled operation.

An organisation may use AI for document processing or forecasting while still relying on disconnected systems, manual data entry and spreadsheets for other critical processes.

📊 Data quality is the biggest obstacle

According to the research, data quality is the leading barrier to scaling AI.

It was cited by 33% of shippers and 45% of LSPs. Integration complexity followed closely, affecting 39% of shippers and 38% of LSPs.

This highlights a fundamental problem with AI adoption in logistics.

AI models can only perform as well as the information they receive.

If shipment data is incomplete, inconsistent or spread across multiple systems, AI has a weaker foundation for forecasting, optimisation and decision-making.

🚚 AI is already producing measurable benefits

Despite the adoption challenges, organisations that are using AI are reporting tangible benefits.

Shippers identified reduced administrative and labour costs as their leading realised benefit, at 55%. For LSPs, improved service was the leading benefit at 53%, followed by freight-cost reduction at 46%.

This suggests that the business case for AI is becoming clearer, even if implementation remains difficult.

🔗 Integration is becoming the real AI challenge

Transportation networks involve carriers, forwarders, shippers, warehouses, ports and other service providers, each generating different types of data.

Connecting those systems is therefore critical.

The Descartes study argues that organisations need the digital infrastructure to connect transportation networks, automate processes and create high-quality, real-time data if they want to move AI from experimentation into scaled operations.
 

💰 Cost pressure is adding urgency

Freight cost management remains the largest current transportation challenge, identified by 35% of shippers and 25% of LSPs.

At the same time, increasing AI adoption is the top transportation priority over the next two years when shippers and LSPs are considered together, at 58%, closely followed by reducing freight costs at 57%.

This creates a dual pressure on logistics organisations:

Use AI to improve operations while simultaneously controlling transportation costs.

🔮 The bigger takeaway

The Descartes findings point to an important reality for the logistics industry.
 

The next AI breakthrough may not come from a more sophisticated model.

It may come from better data, stronger system integration and more automated workflows.

The transformation is moving from:

Manual processes → Digital systems → Connected data → AI automation → AI-driven decisions

For freight forwarders, LSPs and shippers, the message is increasingly clear:

AI readiness starts long before AI deployment.

If the underlying transportation data isn't connected, accurate and accessible, scaling AI becomes significantly harder.