Table 1 Key technological challenges... | Emerald Publishing

Table 1

Key technological challenges from the qualitative data

Specific technological barrier	Key impact	Empirical evidence
Lack of technical AI understanding among professionals	Cannot evaluate suitability or make informed decisions	“Few understand the operational logic of AI applications adequately” (R26)
Unclear security standards and guidelines	Prevents experimentation due to compliance uncertainty	“It has not been jointly agreed whether AI use is secure and data-protected” (R6)
Multiple incompatible information systems	AI solutions cannot connect to existing infrastructure	“several different information systems that do not communicate with each other” (R35)
Budget constraints favor the cheapest solutions	Limits organizational adoption capacity	“only the cheapest solutions are adopted- > resulting in minimal practical benefit” (R16)
Poor and fragmented data availability	Makes AI algorithm training difficult or impossible	“data quality and deficiency” (R56)
Speech recognition and transcription inadequacy	Reduces accuracy and applicability in the Finnish context	“Finnish language speech understanding and transcription” (R58)
Lack of safe testing infrastructure	Prevents proper AI validation before deployment	“lack of testing environments hinders technology adoption” (R22)
Vendor promises vs. real-world implementation	Organizations cannot operationalize promising solutions	“Vendors’ promotional speeches about technology readiness rarely meet the real world” (R37)

Specific technological barrier	Key impact	Empirical evidence
Lack of technical AI understanding among professionals	Cannot evaluate suitability or make informed decisions	“Few understand the operational logic of AI applications adequately” (R26)
Unclear security standards and guidelines	Prevents experimentation due to compliance uncertainty	“It has not been jointly agreed whether AI use is secure and data-protected” (R6)
Multiple incompatible information systems	AI solutions cannot connect to existing infrastructure	“several different information systems that do not communicate with each other” (R35)
Budget constraints favor the cheapest solutions	Limits organizational adoption capacity	“only the cheapest solutions are adopted- > resulting in minimal practical benefit” (R16)
Poor and fragmented data availability	Makes AI algorithm training difficult or impossible	“data quality and deficiency” (R56)
Speech recognition and transcription inadequacy	Reduces accuracy and applicability in the Finnish context	“Finnish language speech understanding and transcription” (R58)
Lack of safe testing infrastructure	Prevents proper AI validation before deployment	“lack of testing environments hinders technology adoption” (R22)
Vendor promises vs. real-world implementation	Organizations cannot operationalize promising solutions	“Vendors’ promotional speeches about technology readiness rarely meet the real world” (R37)

Source(s): Authors’ own work