UniFiji flags AI risks, during Online Safety Act review

The growing use of artificial intelligence (AI) is a concern that requires careful consideration, says University of Fiji vice-chancellor Professor Shaista Shameem.

In a submission to the Online Safety Act review, Professor Shameem said AI was not a significant issue when the Online Safety Act 2018 was introduced, but its influence had expanded considerably in recent years.

“As a university, we’ve got to look at AI and the way it is being used by students and staff,” she said.

“Where is the plagiarism? Where is it useful? How do we make rules to say you can use AI, but only in certain circumstances and not in others? It is now almost impossible to detect AI use.

“It is not like fake news. AI can now interpret facts, and that interpretation may be generated by a person or by a machine.”

Professor Shameem said there were serious questions about how AI operated and how it influenced people's thinking.

“It is a very serious matter now,” she said.

“Anything to do with online communication, not just in terms of safety but also in terms of how we are going to shape our world with this technology becoming such a large part of our reality, is something we cannot avoid.”

She said the Fiji Law Reform Commission was also examining the issue as part of efforts to strengthen the Act for the future.

“We found many weaknesses when we reviewed the 2018 Act,” Professor Shameem said.

“There were weaknesses in the structure and in the Government’s model, including whether there should be one commissioner or several commissioners.

“We favour several commissioners because it avoids conflicts of interest.

“We also believe the structure should more closely mirror the Human Rights and Anti-Discrimination Commission because it has not only an education function but also robust enforcement provisions.”