Report warns of regulatory risks telcos face when adopting AI

The report – titled Regulatory Risks and Challenges of Adopting Artificial Intelligence (AI) in the Telecoms Sector – points to ‘critical areas’ it believes telcos must focus on with regards to AI regulation, which include rules surrounding high risk situations and prohibited use, as well as transparency requirements, and enforcement.

“While the use of AI creates many benefits such as improved customer service, optimized network operations, detecting new business opportunities, cost savings, and process automation, there are also numerous risks and challenges to be aware of”, said Sarah McBride, Principal Analyst, Regulation at Omdia.

These risks also include availability of unbiased and reliable data, liability in case of error, lack of trust in decision making, privacy and security, says the report.

It also says while most AI systems in telecoms are still under development, AI is poised to play a ‘significant role in the sector in the mid-term’, and therefore it is ‘essential’ that telcos are fully up to speed on the potential regulatory risks and challenges that come with AI adoption.

“There has been a surge of activity in the AI space, particularly around potential regulation, and this trend is likely to continue”, McBride said. “As a result, it would be prudent for telcos to proactively implement a risk management framework within their systems before AI implementation, ensuring adherence to standards of data quality, accuracy, robustness, and non-discrimination. Equally important is the adoption of an ethical, human-centric and trustworthy approach to AI. As key enablers of AI technology, it is crucial for the telecoms sector to drive ethical and responsible AI technology.”

Telcos will need to adhere to different guidelines depending on the context and application of AI, says the report, pointing to existing regulatory tools in the EU such as the Data Act, the Digital Markets Act, and the AI Act, as well as regulations not specifically targeting AI such as the European Electronic Communications Code, net neutrality, data protection and privacy regulations.

The reports adds that AI technologies that pose risks to health, safety or rights particularly those involving biometric information, network security or critical infrastructure are subject to heightened scrutiny with regards to adherence to data protection, security and privacy of the large data sets that underpin them.

“Telcos must ensure they fulfil regulatory requirements for these high-risk situations such as transparency, human oversight and data governance by completing risk assessments and implementing governance frameworks”, said McBride.

A recent Telecoms.com survey showed that operator’s comfort limits regarding the adoption of AI are reached with AI chatbots and ‘supervised’ automation in a 5G network. The report says: “It is revealing the extent to which telcos feel most comfortable adopting AI and automation in a closed loop manner and handing over the reins to these technologies in their production environments.”