- Operators like Orange Business and AT&T are bullish on AI, despite talk of a bubble
- Successful operators focus on a disciplined approach based on prioritizing ROI, business value and controlling data
- AI "Pacesetters" that maintain business discipline "outperform their peers across every measure of AI value," according to a Cisco report.
There's a storm coming and it's headed for telcos. OpenAI CEO Sam Altman warned about an AI bubble back in August, a caution that was followed by warnings from the Bank of England, International Monetary Fund and Amazon founder Jeff Bezos.
Nvidia CEO Jensen Huang, White House AI czar and venture capitalist David Sacks and AMD CEO Lisa Su have said nope no bubble here. Analysts, however, have kept up the warnings, arguing that investments in AI grossly exceed returns. A Bank of America analyst claimed we're in an AI "air pocket," not a bubble. Whatever that means.
Given the risk of a bubble and its potential collapse, how are communications service providers looking ahead? How can telcos not just survive the coming AI storm but thrive?
Fierce spoke with leading telecom executives and industry analysts about how successful telecom companies should navigate the AI storm. Rather than being hunkered down in subterranean survival bunkers, stocked up on beans, bottled water, gold and bitcoin, prepared to weather the imminent apocalypse, these leaders were just...working. Going about their business.
And that's the prescription for how telcos should leverage AI: Forget about all the talk about a bubble. Treat AI like any other emerging technology. Invest wisely, use the technology to achieve measurable business results, involve the whole business in planning and execution, measure outcomes and, based on those measurements, increase investment in what's working and adjust course or abandon what failed.
Inflated short-term AI expectations, real long-term value
"The benefit that AI brings to business will be enormous," said Usman Javaid, Orange Business chief product and marketing officer. "While short-term expectations might be inflated, the long-term value that AI will deliver is undeniable."
For the short term, "Many companies are still in the proof-of-concept phase, and we see reports out of MIT that the majority have yet to see significant financial returns from their AI investments," Javaid said. He was referring to a July MIT study, "The State of AI in Business 2025," which found that 95% of organizations are getting zero return on AI, despite $30-$40 billion in investments.
Nonetheless, Javaid is bullish on AI. "This is one of a few lifelines left for the telco industry," he told Fierce in May. Telcos can succeed by providing digital infrastructure to enterprises that are adopting AI.
Orange Business is providing connectivity; a private cloud with sovereign AI services across Europe; a platform called Live Intelligence to help enterprises of different sizes build generative AI systems; and an AI-driven contact center solution in partnership with Microsoft.
The AI hype cycle is peaking
Generative AI and agentic AI are at the peak of the famous Gartner hype cycle. That is, they're overhyped, Gartner Senior Director Analyst Susan Welsh de Grimaldo told Fierce Network. A natural market correction will come. But AI will deliver long-term results.
"If there is an AI bubble and it bursts, AI still has transformational potential and that does not go away," de Grimaldo said. "We will still be in an AI race and the race will be about delivering real value."
The companies that prosper will be the ones that show real differentiation. "It will be a marathon," she said. The market correction will separate the pack leaders from followers.
"When the dotcom bubble burst, the Internet did not go away," de Grimaldo said. "It became foundational to many things and everyone uses it in their daily lives."
Orange Business's Javaid agreed. "Some level of market correction or volatility is likely coming, and the AI or tech bubble might see setbacks along the way. But in the bigger picture, the value and potential of AI to reshape industries will still be there. The short-term ups and downs are part of the natural cycle of innovation," he said.
"Even if the overall tech bubble cools off a bit, the business-specific bubbles that are driven by real needs will continue to grow," Javaid said.
He added that Orange Business is taking a cautious and steady approach, with a focus on building flexible and sovereign platforms with a variety of partners. That way, he explained, it can provide choice and the ability to better protect client data and investments if the market shifts.
As it lays its AI foundation, Orange Business is prioritizing strong data governance, security and ethical AI practices, Javaid said.
He continued, "We’re also tackling some of the fundamental challenges, such as fragmented IT systems and data silos."
Succeeding with a grounded approach
AT&T is similarly maintaining a disciplined approach to AI investment. This philosophy has helped the company deliver a 2x free cash flow ROI on every dollar invested in AI, said Andy Markus, AT&T chief data and AI officer.
"We're leveraging the reality of AI today to do compelling things for AT&T, and whether there's a bubble or not from a stock perspective, it doesn't impact our ability to drive significant, meaningful value from the technology as it stands today," Markus said.
AT&T's successful AI initiatives include Ask AT&T, its generative AI platform for customer support, along with the new Ask AT&T Workflows for agentic AI. These platforms have 100,000 users across AT&T, with 750 million API calls into production systems. AT&T also uses AI for software development, supporting technicians in the field and back office.
"It touches everybody and almost everything," Markus said. "The scale really matters." AT&T focuses on accuracy. "We can deliver accuracy beyond the level that most companies can," Markus said.
Data, he added, is the essential foundation on which to build a successful AI strategy.
"Control your data. It's the data that drives the value," Markus said. "We don't care if generative AI can write a poem — we care if it's accurate on AT&T information."
An Accenture study "Cracking the Code on Data Debt," found that “data debt”, or fragmented and inconsistent data, is preventing many communications service providers from leading in the AI economy. It also shows that data-driven, AI-first CSPs are more adept at using AI to strengthen customer loyalty, lower customer acquisition costs, launch products faster and resolve network faults more quickly.
"Telcos need to be very careful and not get ahead of themselves," said analyst Roy Chua, founder and principal at AvidThink. "The good news for telcos is a lot of the senior telco management were there for the 2000 dotcom crash, followed by the telco crash. They will be a little more cautious."
A recent Cisco survey sheds a favorable light on companies — like AT&T and Orange — that adopt pragmatic AI strategies. The Cisco AI Readiness Index, now in its third year, identifies a small group of 13-14% of companies, which it calls "Pacesetters," that "outperform their peers across every measure of AI value." These companies are "3X more likely to track and measure the impact of AI investments" and "1.5x more likely to report gains in profitability, productivity and innovation."
The lessons from leading telcos and analysts are clear: Treat AI like a long-term investment, not a gold rush. Prioritize accuracy and measurable ROI, move proven use cases from PoC to production and control your data.
That way, when the AI bubble collapses, it will be just another day delivering business value.