San Francisco, Feb 8 More than one-in-ten workers will try to trick artificial intelligence (AI) systems used to measure employee behaviour and productivity by 2023, according to a prediction by Gartner on Monday.
Such systems have seen a major rise in use in the wake of the Covid-19 pandemic.
"Just as we've seen with every technology aimed at restricting its users, workers will quickly discover the gaps in AI-based surveillance strategies," Whit Andrews, Distinguished Research Vice President at Gartner, said in a statement.
"They may do so for a variety of reasons, such as in the interest of lower workloads, better pay or simply spite. Some may even see tricking AI-based monitoring tools as more of a game to be won than disrespecting a metric that management has a right to know."
Organisations are using AI-enabled systems to analyse worker behaviour in the same way that AI is used to understand shoppers, customers, and members of the public.
These tools provide basic activity logging with alerts, or in more sophisticated versions, can attempt to detect positive actions or misbehaviour through multivariable analysis.
Many employers use productivity monitoring systems despite a high percentage of workers finding such tools unappealing.
Even prior to the pandemic, Gartner research showed that workers feared new technologies used to track and monitor work habits.
As these tools become more prevalent, organisations will increasingly face workers who seek to evade and overwhelm them, Gartner said.
Workers may seek out gaps where metrics do not capture activity, accountability is unclear, or the AI can be fooled by generating false or confusing data.
"IT leaders who are considering deploying AI-enabled productivity monitoring tools should take a close look at the data sources, user experience design and the initial use case intended for these tools before investing," said Andrews.
( With inputs from IANS )
Disclaimer: This post has been auto-published from an agency feed without any modifications to the text and has not been reviewed by an editor