• Survey commissioned by Ciena found 90% of service providers believe AI demand will drive revenue growth from datacentre connectivity services 
• Managed optical fibre network (MOFN) services expected to play key role in delivering AI to enterprises
• Many operators believe optical network upgrades will be key to maximising the new business opportunities

The majority of wide area data network operators expect high-capacity connectivity to and from datacentres, driven by the increasing use of AI, will be a primary driver of revenue growth in the coming years, but that many also recognise an urgent need to upgrade their optical transport networks in order to meet that demand, according to a new report commissioned by Ciena.

Research carried out by Censuswide on behalf of the optical network equipment vendor found that 90% of the 1,200 executives at telecom operators, wholesale fibre network operators and regional service providers surveyed believe that demand from enterprises, hyperscalers, and neoscalers (aka neoclouds) for connectivity to and from AI infrastructure facilities will be a key factor in driving revenue growth during the next three to five years, with 56% citing them as their primary source of net-new revenue. 

The survey, which gathered feedback from network operators across 12 countries, also found that 96% of respondents expect managed optical fibre network (MOFN) services to generate revenue from connecting distributed AI clusters, while 51% identified MOFN as their primary vehicle for meeting datacentre interconnect (DCI) demand.

Analysing the services in which operators had most confidence as a source of revenue, AI inferencing was selected by 49%, while 94% of respondents said they see significant opportunities to capture GPU-as-a-service revenue through essential strategic partnerships with cloud and neocloud providers. In addition, 95% of respondents believe immersive and AI-driven entertainment will also contribute new revenue streams in the next three years, though just 29% named wearables and personal devices – such as smart glasses and health trackers – as the leading consumer hardware category for driving premium service revenue. 

Physical AI, which is especially being advanced in China, is also seen as a major opportunity, with 60% of respondents saying they expect physical AI ecosystems, such as industrial robotics, to account for more than 15% of their enterprise AI revenue within five years.

Despite the positive feeling about the opportunities offered by AI, the service providers polled also expressed a sense of urgency for the need to upgrade optical networks to meet the expected demand from enterprise customers (which, of course, is music to Ciena’s ears…). 

Nearly half (48%) believe those upgrades are critical, and 39% said they are needed within the next 12-18 months. Just 11% believe that routine upgrades will be able to meet expected demand, and additional network automation is seen as essential, according to 96%.

“The survey findings show that service providers are approaching AI with both optimism and urgency," said Brodie Gage, Chief Product and Technology Officer at Ciena. "Service providers see significant opportunities around AI inference, GPU-as-a-Service, high-capacity network services, and quantum-safe encryption, while recognizing that many existing networks must evolve to meet new performance demands. The message is clear: AI is creating new sources of revenue, and success will depend on having the network foundation to support them.”

Censuswide spoke to 100 executives between July 13 and July 23, 2026 in each of the following 12 countries: Australia, Brazil, Canada, India, Indonesia, Saudi Arabia, Mexico, Singapore, UAE, UK, US, and Vietnam.

Looking at a breakdown by country, the study found that 92% in the US agree that high-capacity AI-driven services will be a primary driver of net-new revenue growth, compared with just 61% in Canada, with Brazil (64%) and India (67%) also showing significantly less confidence. 93% of Brazilian respondents pointed to the importance of premium, guaranteed low latency SLAs (service level agreements) for monetising enterprise edge deployments.

Service providers in Australia put significantly less weight on multi-cloud connectivity management as a revenue engine (38%) than the global benchmark (56%).

100% of respondents in the UAE considered advanced network automation as crucial or important to capturing AI-driven opportunities.

- James Pearce, Editor, TelecomTV