How AI and Disinformation Will Affect Future Emergencies
Tracks
Prince Room | In-Person & Virtual via OnAIR
| Tuesday, July 23, 2024 |
| 10:50 AM - 11:10 AM |
Overview
Ryan McBeth, Thomas Green & Ilana Pender-Rose, Independant
Details
Key Presentation Learnings:
1. Understanding what Deceptive Imagery Persuasion (DIP) is within the context of an emergency;
2. How a media and communication strategy could be designed to mitigate DIP from impairing or degrading communications during an emergency; and,
3. How regulatory amendments could be made to reduce the likelihood of a DIP-attack from occurring during the course of an emergency.
Speaker
Mr Thomas Green
Emerging Technologies
Independent
AI and Emergency Communication
Biography
Thomas Green specialises in regulation and implementation of new technology. He is a co-founder of Neumann Space, and has published extensively on issues concerning access to space by commercial and non-government entities. He recently launched an award-nominated, Australian-first trial of augmented enforcement signage to improve road and rail safety across NSW; and has implemented trial evaluation for space-based technology to improve asset management, transport resilience and business continuity for NSW.
Mr Ryan McBeth
Director of Integration & Intelligence Analyst
University of Texas
AI and Emergency Communication
Biography
Ryan McBeth is an intelligence analyst, software architect, cybersecurity guy, and YouTube Content Creator of Military and Intelligence Topics. He appears on NEWSMAX as an intelligence consultant and consults for Veloxxity on OSINT and Information Warfare.
He also works for the Texas AeroMedLab, University of Texas at Dallas as the Director of Integration. This program develops AI powered drones that autonomously deliver whole blood to wounded soldiers on the battlefield.
Ryan spent 20 years as an anti-armor and heavy weapons infantryman with two overseas deployments and developed C4ISR software for Accenture Federal Services.
Ms Ilana Pender-Rose
Independant
Independant
AI and Emergency Communication
Abstract
The community is at its most vulnerable of undergoing a mass casualty event; or, significant loss of property, during an emergency.
The ability for a prescribed combat agency to effectively communicate notifications, warnings and evacuations to a community through new-media channels may be degraded where interference by third parties occurs. Such interferences may impair a combat agency’s competency to optimise community resilience, business continuity and public trust; and foreseeably increase the scope of loss of life, property and environment caused during a disaster; as well as greatly delay the recovery phase.
With AI-generated deep-fake videos, photos and audio becoming ubiquitous, alongside AI-generated text, the line between what is real and what is fake has become blurred. The emergence of GTP+3 and higher text and visual AI can generate convincing videos of public figures, fabricated images of events; and author news stories void of any factual basis within seconds.
Such AI-Generated materials may be used by third party actors to undertaking Deceptive Imagery Persuasion (“DIP”) - the method of using unrelated materials and passing them off as related to an existing event. Through utilising AI generated materials for the purposes of DIP and interfering with community notifications, an actor may foreseeably increase the risk of casualties, destruction of property or environment during the course of an emergency.
This presentation will unpack the risk of AI-Generated material being utilised for the purposes of DIP, and provide recommendations for express legislative reform to prohibit DIP activities being undertaken in relation to a community event. In addition to this, this presentation will also recommend ways of mitigating the risks of DIP activities impairing on emergency management and response.
The ability for a prescribed combat agency to effectively communicate notifications, warnings and evacuations to a community through new-media channels may be degraded where interference by third parties occurs. Such interferences may impair a combat agency’s competency to optimise community resilience, business continuity and public trust; and foreseeably increase the scope of loss of life, property and environment caused during a disaster; as well as greatly delay the recovery phase.
With AI-generated deep-fake videos, photos and audio becoming ubiquitous, alongside AI-generated text, the line between what is real and what is fake has become blurred. The emergence of GTP+3 and higher text and visual AI can generate convincing videos of public figures, fabricated images of events; and author news stories void of any factual basis within seconds.
Such AI-Generated materials may be used by third party actors to undertaking Deceptive Imagery Persuasion (“DIP”) - the method of using unrelated materials and passing them off as related to an existing event. Through utilising AI generated materials for the purposes of DIP and interfering with community notifications, an actor may foreseeably increase the risk of casualties, destruction of property or environment during the course of an emergency.
This presentation will unpack the risk of AI-Generated material being utilised for the purposes of DIP, and provide recommendations for express legislative reform to prohibit DIP activities being undertaken in relation to a community event. In addition to this, this presentation will also recommend ways of mitigating the risks of DIP activities impairing on emergency management and response.
Biography
Ilana Pender-Rose is a media professional with fifteen years’ experience in media and emergency management. Ilana has worked in radio, digital media and information communication technology as well as print and online publications. She has covered emergencies working both in the media industry and without, including the Hawkesbury Nepean Floods, COVID-19; the Martin Place Siege and the Black Summer bushfire for which she received the NSW Premier’s citation.
