- DT just inked a deal with OpenAI to co-create internal and customer-facing AI products
- DT is one of few operators taking a full stack approach but its strategy could be hard to replicate
- OpenAI appears to be cultivating telcos as a key channel to get its AI to a broader audience
Anyone who does puzzles knows there’s something infinitely satisfying about putting that final piece into place. It seems that’s exactly what Deutsche Telekom (DT) has done with its new OpenAI collaboration.
The operator just inked a multi-year deal to work with OpenAI on both internal and customer-facing AI applications. On the latter front, the pair are set to work on “simple, personal and multi-lingual AI experiences” for both communication and productivity, with pilots planned for Q1 2026.
Internally, DT will rollout ChatGPT Enterprise across the company, targeting increased use of AI for customer care, internal copilots and network operations. Tekonyx President and Chief Research Officer Sid Nag told Fierce this could take the form of more AI customer service assistants that can help with billing questions, service outages, plan changes, roaming, device troubleshooting; personalized plan recommendations; or proactive network quality notifications.
DT might also use ChatGPT-based coding tools “to build new applications that are used for DT network provisioning, billing such as OSS, BSS systems and tools,” Nag added.
A cut above the rest
Futurum Group VP and Practice Lead Nick Patience told Fierce the arrangement is kind of a big deal for DT.
“This deal elevates DT from being a user of AI to being a co-developer, which is pretty significant,” he told Fierce. “DT is one of the few operators building a full-stack AI story.”
As Patience noted, its work with OpenAI is the latest play in DT’s years-long quest to position itself as a leading light on cloud and AI in Europe. The operator has been all-in on digital sovereignty since 2021, when its T-Systems arm teamed with Google Cloud on sovereign cloud services.
Earlier this year, DT also launched a new suite of services under the T Cloud brand, offering sovereign public, private and AI cloud options using a combination of hyperscaler and its own cloud infrastructure. Just two months later, it followed up by announcing a collaboration with Nvidia to build an industrial AI cloud (in the form of a data center in Munich) for Germany, which is scheduled to begin operations in Q1 2026.
During an earnings call in November, DT CEO Timotheus Hoettges noted the operator is also in the running to deploy one of five AI gigafactories the EU is aiming to build as part of its 200-billion-euro InvestAI initiative. Around 20 billion Euros of the funding will go toward the gigafactories, which will run 100,000 GPUs each.
Like Patience, Recon Analytics Founder Roger Entner told Fierce that the addition of OpenAI to the mix means DT can now “legitimately go to their clients and can say ‘hey, we have the connectivity, data center and the software” to run their workloads – all wrapped up with a sovereign bow.
Nag added OpenAI’s tech could help DT generate new revenue from offerings like edge AI compute for enterprises, vertical AI services for industries like healthcare, retail and manufacturing, and private 5G + AI bundles for factories, campuses and other logistical hubs.
A model worth following
Asked whether DT’s strategy is a model for other operators, Entner said the answer depends on where those operators are. For most European incumbents like France’s Orange or Spain’s Telefonica, the answer is yes. For U.S. operators, not so much.
“There are not that many AI data centers in Europe and in Germany,” Entner explained, noting this leaves the door open for operators like DT to fill in the gap. “In the U.S. you have a ton of data centers that you can do AI. Therefore, it doesn’t make sense for a network operator to have also a data center. They tried to compete with hyperscalers, and it failed. And the scale in the U.S. is a lot bigger than in Europe.”
As for operators in the U.K., Entner said it’s “little bit late” given the country has just announced plans to build a slew of AI data centers. That, he noted, will create a situation similar to what U.S. operators are facing.
AvidThink Founder and Principal Roy Chua similarly said that DT's "reference blueprint" of partnering with a hyperscaler, Nvidia and OpenAI to cover all layers of the stack might not be something that other operators can simply copy.
"DT has scale benefits, including 261 [million] mobile subscribers across Germany and other European subsidiaries, as well as deep pockets for billion-euro investments. Plus, it has the incentive of European regulatory compliance around data sovereignty and government backing," he told Fierce. "There’s still significant execution and go-to-market risk, and we haven’t yet seen clear wins across CSPs in AI initiatives."
Beyond hyperscalers
The deal is also the third collaboration between OpenAI and a major telecom operator, following its work with DT subsidiary T-Mobile and SK Telecom.
Nag noted there is clearly a strategic motive behind these tie-ups.
“Telcos – if they execute – will have a big play in the edge inferencing space as well as providing hosting and colo services that can host domain specific SLMs that need to be run closer to the user data,” he said. “Furthermore, telcos will play a role in connectivity services across Neocloud providers such as CoreWeave, Lambda Labs, Digital Ocean, Vast.AI etc. OpenAI does not want to lose the opportunity to partner with telcos so they are striking early.”
Patience agreed.
“This is an example of OpenAI treating telcos as high-scale distribution and data channels - customer care, billing, network telemetry, national reach and government relationships,” he concluded. “This suggests OpenAI is deliberately building an operator channel in key regions (U.S., Korea, EU) but still in partnership with existing cloud and infra providers rather than displacing them.”
Update 12/9/2025 3:20 pm ET: This story has been updated to include comments from Roy Chua.