28
AI &AUTOMATION BAROMETER REPORT
• Not missing out, but also not investing in things that will turn out not to be useful
• Pricing
• That clients would not want to pay for it, having an illusion that they can do it all by themselves...
• Time and resources efficiency
• Trust. And the fact that clients think they can now manage the translation process on their own
without the help from LSPs.
Data &security concerns
• Confidentiality (x3)
• Data security (x3)
• Ethical adoption with accountability -need ways to measure as they do not exist today for
interpreting.
• Ethical usage
• Legal issues
• One of our concerns as early adopters of AI related to security of information, however, we have
found our current implementation model to provide a private and secure environment that does not
put the data at risk.
• Privacy (x4)
• Security (x3)
Quality and technical issues
• AI behaviour
• Control (x2)
• Disruption of current workflows
• Ease of identification, implementation and getting expected output. Despite being an advance tool of
the era, there’s lot of confusion on user side.
• Insufficient size and quality of training data.
• Integrations
• Lack of integrations with existing tech stack
• Providing faulty results to customers
• Quality of the output (x2)
• Reliability
CHALLENGES
AI &AUTOMATION BAROMETER REPORT
• Not missing out, but also not investing in things that will turn out not to be useful
• Pricing
• That clients would not want to pay for it, having an illusion that they can do it all by themselves...
• Time and resources efficiency
• Trust. And the fact that clients think they can now manage the translation process on their own
without the help from LSPs.
Data &security concerns
• Confidentiality (x3)
• Data security (x3)
• Ethical adoption with accountability -need ways to measure as they do not exist today for
interpreting.
• Ethical usage
• Legal issues
• One of our concerns as early adopters of AI related to security of information, however, we have
found our current implementation model to provide a private and secure environment that does not
put the data at risk.
• Privacy (x4)
• Security (x3)
Quality and technical issues
• AI behaviour
• Control (x2)
• Disruption of current workflows
• Ease of identification, implementation and getting expected output. Despite being an advance tool of
the era, there’s lot of confusion on user side.
• Insufficient size and quality of training data.
• Integrations
• Lack of integrations with existing tech stack
• Providing faulty results to customers
• Quality of the output (x2)
• Reliability
CHALLENGES
