Comms industry confronts AI's data quality problem

Communications professionals and data firms are converging around a shared concern in the AI era. Put simply, artificial intelligence systems are only as trustworthy as the information fed into them. 

At the AMEC Global Summit in Dublin this week, discussions repeatedly returned to questions of data quality and governance, as communications leaders grappled with the practical limitations of deploying AI tools.

During a session led by representatives from the European Commission, speakers described AI implementation as dependent on a “hierarchy of needs”, arguing that organisations cannot build effective AI systems without first establishing reliable and well-structured underlying data.

One of the main interests among attendees was the “semantic layer” underlining communications data, referring to the contextual architecture allowing AI systems to interpret meaning rather than simply process information.

Implementation costs and the challenge of securing long-term organisational buy-in for large-scale data infrastructure projects were further concerns. The discussion suggested a growing recognition across the communications industry that many organisations remain structurally unprepared for the AI systems. 

The summit has also revealed how the communications ecosystem itself is developing in response to AI-driven discovery systems, or “generative experience optimisation” (GEO). Speaking to Communicate, Victoria O’Brien, CMO at PR firm 72Point, said that newswire and content supply firms were becoming increasingly important as publishers face pressure to produce high volumes of search-friendly content. 

Across the event, conversations suggested that the scope of communications professionals is moving beyond traditional media advisory and towards shaping the infrastructure through which information is organised and trusted online.