Lumen CEO says AI bots are taking over the Internet

OVER half of the planet’s Internet traffic is now made up of AI bots, according to Kate Johnson, chief executive officer of enterprise network giant Lumen Technologies, forcing executives across sectors to rethink how their companies handle everything from customer-service requests to hidden network threats.  

On Monday, Johnson penned an open letter to fellow company heads, warning that they must prepare for the seismic shift in AI-driven traffic patterns at volumes and speeds that are harder to predict. 

“The intensity of, and pace and the volume of data is proliferating quite rapidly,” she said in an interview. “More than 50 per cent of the traffic on the Internet today is created by autonomous workers. That’s remarkable because as most CEOs would tell you, we are just beginning our AI journey. So if it already comprises 50 per cent, imagine what it’s going to look like in a year, three years, five years.” 

Companies that rely on computing power and data processing, Johnson said, can no longer take their physical network capabilities for granted.

Organisations that used to purchase network capacity to cover set volumes and routes have to rethink those static patterns and consider switching to a consumption-based model.

Johnson took the helm at Lumen, which handles 65 per cent of global Internet traffic, in November 2022, the same month that OpenAI’s ChatGPT was first released. Since then, she’s been reorienting Lumen to adapt to the AI boom.

Under Johnson, the company has sold its consumer fiber-to-the-home business to AT&T for US$5.75 billion and used the proceeds to pay down US$4.8 billion in debt. She is chasing US$1 billion in cost savings by the end of 2027 and has forged a US$200 million software partnership with Palantir Technologies.

The relationship gives Lumen better visibility into the scope and needs of its customer base, Johnson said, as well as suggesting “what our next logical business opportunity is.” 

Lumen is incorporating AI into its own operations, using multiple large language models to help identify network threats.

SEE ALSO

OpenAI’s US$852 billion valuation faces investor scrutiny amid strategy shift

AI is on its way to upending cybersecurity

The coming AI-driven ‘abundance’ shock

The company said that in late 2024 it detected traces of activity tied to Salt Typhoon, a Chinese hacking group targeting US telecommunications networks, and subsequently ejected the suspected perpetrators. AI capabilities, like those now on display in Anthropic PBC’s powerful new Mythos model, will only improve such responses, she said.

“The way that you detect threats and the way that you thwart them is through observation,” Johnson said. “The threat actor can actually collect intelligence about vulnerabilities in record time, which means that you have to have equal or better AI capabilities to thwart their intelligence.” BLOOMBERG