Britain’s telecommunications regulator has issued a warning about the potential risks associated with the increasing use of artificial intelligence (AI) to manage phone and broadband networks. Ofcom raised concerns that the automation of network operations could lead to “catastrophic” failures, urging major operators such as BT, Vodafone, and Virgin Media O2 to proceed with caution.

Telecom companies have been adopting AI technologies to detect and resolve network issues more swiftly, which can improve service reliability. However, Ofcom emphasized that automating network management carries inherent risks, particularly due to the possibility of unexpected system malfunctions or vulnerabilities being exploited.

The regulator’s alert highlights broader security worries around AI, especially “frontier” models that could facilitate hacker access to critical infrastructure. Earlier this month, BT became the first UK company to participate in a trial of Anthropic’s Claude Mythos Preview, an advanced AI model considered too sensitive for public release. The trial was subsequently suspended after former US President Donald Trump intervened to restrict foreign nationals from accessing the technology, underscoring geopolitical complexities surrounding AI tools.

Ofcom’s new guidance reflects a cautious stance on deploying AI in telecommunications, stressing the need for careful assessment and governance at every stage of automation implementation to mitigate risks.

Responding to the concerns, a BT spokesperson said the company has “comprehensive controls and governance in place” for the use of automation across its networks. The statement added that BT continues to evaluate emerging technologies to enhance both network performance during critical failure scenarios and overall cybersecurity.

As telecom providers increasingly integrate AI into their systems, balancing innovation with operational security remains a key focus for regulators and industry stakeholders alike.