Senate Finance and Public Administration Legislation Committee, , 4:05:36 to 4:07:23
NIAA executives discuss the proliferation of AI-generated complaints, which increase the administrative burden on integrity teams by requiring more time to analyze lengthy, automated text for genuine issues.
Topics: First Nations, Artificial intelligence, Integrity
Speakers: Speaker 22, Julie-Ann Guivarra, Senator Kerrynne Liddle, Unknown Speaker
>> and senator it was something that we actually discussed at our executive board just last week and one of the other trends that we are noticing is actually just the proliferation of complaints generated by AI as well. It was [clears throat] a conversation that we had around um how and this is not a uncommon issue across other public service agencies as well is just um responding in a timely manner to what is uh you know an increase in the number of complaints that are put that that are being seen across complaint mechanisms in other agencies as well largely generated by artificial intelligence.
can you tell me how many of those actually are? I mean, they're they're pretty obvious if they're form letters, if they're complain, if they're AI generated. I mean it's a big claim to say that um AI is playing a part in
>> senator by by their nature um and and I I don't want to um you know by no means raw complaints being AI generated but there are some clear markers for those that that are clearly AI generated instead of being a perhaps a one or a two-page letter of complaint we could be receiving up to 15 or 16 pages of of AI generated um text that that uh often brings in matters quite unrelated to to the nature of the the complaint itself. Uh it is resulting in additional effort and time from the complaints team to actually analyze these letters and really understand the nature of the complaint. um understand who the appropriate party is that that we can deal with so that we can as I said before ensure that we are dealing with complaints as quickly as we can and as appropriately as we can.