Governments and businesses have been warned that it may not be possible to control or mine what AI does with people’s personal data once it has it.
Photo: 123 RF
Warnings have been published by the Privacy Commissioner about keeping people’s data safe when using artificial intelligence, which applies to both the private and public sectors.
Privacy Commissioner Michael Webster issued the prospectus on Thursday, warning that as technology changes rapidly, the advice will evolve to keep up with the changes.
He said that any organization that uses artificial intelligence must handle people’s personal information carefully.
While the technology has been chosen to lead to significant productivity gains, it also carries significant privacy risks.
Data input to AI such as ChatGPT is non-retrievable, and there are limited controls over its use.
“I expect agencies to do due diligence and privacy analysis to assess how they comply with the law before they jump in to use artificial AI,” he said.
He warned against entering personal or confidential information into the technology unless there was “clear justification”.
“An alternative would be to strip the input data of any information that would allow re-identification.”
Given the potential privacy implications, employees should consider whether it is necessary and reasonable to use AI, or if there is a different approach.
Employees must sign off from supervisors and privacy officers and tell customers when AI is being used.
According to Webster, a person should investigate any information generated by the technology before taking any action because of it.
It would expect agencies to conduct due diligence and privacy analysis, including a privacy impact assessment, before using the technology.
They should: “seek feedback from affected communities and groups, including Māori” and “ask the provider to explain information and evidence about how privacy protections have been designed into the system.”
The commissioner has previously sent a letter to government agencies warning of the need to use AI without proper assessment and a government-wide response to the growing challenges posed by the tool.
[ad_2]