Persicaria: Promoting ethical AI literacy into mainstream societal consensus

Our team wants to continue making educational resources that improve ethical AI literacy in general society. We hope to do this through making interactive content and hosting both online and in person workshops, hoping to translate the concept of responsible usage into a more understandable appeal to the everyday person. We currently have a team of around ~30 people who are spread around making the resources, researching the risks, gathering user feedback on their perceptions of potential issues and the delivery team. We believe that currently most of the general public is unaware of the potential risks of AI misuse, with most day to day usage being self centred. As such, our material tutorials will focus on raising awareness on preventable adverse impacts of AI usage, and hope to cover a wide range of fields from its implications on academic integrity and malicious usage for harassment in schools, to its risks in company and corporate decisions such as making policy or financial decisions based on findings on hallucinated data, or privacy breaches from carelessly inputting sensitive data such as personal records into LLMs. The output is that we hope to see a trend of negligent AI usage decrease across all fields, and achieve make responsible AI usage mainstream societal consensus based on a thorough understanding of risks rather than personal intitution on what is right or wrong.

Currently most of the safeguards against adverse AI implications are focused on the hardware side, and we are worried about catastrophes simply stemming from user end negligence. We believe that encouraging ethical usage is a very cost effective way to mitigate harm in any future scenario.  We seek the advancement of AI literacy in becoming an accessible, fundamental right imperative in avoiding adverse risks that arise from the exponentially growing gap between AI capability and a global populace mostly underequipped with harnessing its potential for benefit. No matter how robust a technical safety mechanism could be, the capacity for all of that to be undone by simple user negligence is something that is painfully overlooked. This has already become a reality and could progress to something much worse- for instance, where we are based in, Australia, saw consulting companies give an AI generated report with hallucinated findings that would have influenced national policy and decision making, costing taxpayers over 500k, and also would have carried out more serious implications if critical decisions were made simply to due a lack of ethical consideration when utilising AI tools.
The incident: (https://www.theguardian.com/australia-news/2025/oct/06/deloitte-to-pay-money-back-to-albanese-government-after-using-ai-in-440000-report)

If down the line, such sheer negligence had been failed to be identified and thus applied onto the decision making of governmental bodies, it could lead to policies that dramatically alter the livelihoods of entire societies, and already can bring harm before AGI would be even achieved. Even individually, a sole drone operator in war who fails to check on whether a target identified by AI is a civilian or a legitimate combatant could carry immense moral ramifications.

The funding will be spent on enabling our team to cover more specific fields of end-user risk (corporate environments, decision making with AI) and also enable us to keep up with the potential risks from newest emerging models or technologies, and also extend our workshops to be hosted more frequently as it is currently done by volunteers taking their own time. We can also fully commit towards producing much more quality, up to date lessons responsive in addressing aforementioned evolving AI capabilities,  and hopefully connect with like minded educators that can giving us a platform to provide accessible resources globally.