Let’s Chat about Gross Public Text generation
ChatGPT is the buzzword of the year. Suddenly everyone mentions and everything uses ChatGPT (or so they say). But many people do not know what ChatGPT really stands for, or what it really is! Many people do not even know ChatGPT is just an example of a Large Language Model (LLM) bundled with some Machine Learning (ML) module and that there are many others in existence. To get a good overview of the progress in LLM technology, we will dig into the short history of LLMs.
There is no doubt about that LLMs can be a great asset for your company’s workflow. But at the same time, you should wonder: can an LLM expose your company secrets? The answer is plain and simple: YES! (And it already happened.)
Leaving aside the good things an LLM can be for your company, we will focus on the dangers lurking around the use of an LLM and/or AI and your company depending on it, and showing real-life examples where it went or can go wrong, e.g., with Identity Management, Marketing, PII, (False) Information gathering, creation of (unethical) content, etc.
Of course, LLM and ML failures can lead to some humorous moments too; several of these will be showcased during the presentation.
We will conclude with what you can do to safeguard your company and yourself against all the nasty elements that may occur when using LLMs.
Mr. Righard Zwienenberg
Zwienenberg started dealing with computer viruses in 1988 after encountering the first virus problems at the Technical University of Delft. His interest thus kindled and studied virus behavior and presented solutions and detection schemes ever since. Initially starting as an independent consultant, in 1991 he co-founded CSE Ltd. In November 1995 Zwienenberg joined the Research and Development department of ThunderBYTE. In 1998 he joined the Norman Development team to work on the scanner engine. In 2005 Zwienenberg took the role of Chief Research Officer. After AMTSO – Anti Malware Testing Standards Organization – was formed, Zwienenberg was elected as president. He is serving on the board of AVAR and on the Technical Overview Board of the WildList. In 2011 Zwienenberg was looking for new opportunities and started as a Senior Research Fellow at ESET. In April 2012 Zwienenberg stepped down as President of AMTSO to take the role as CTO and later as CEO. In 2016 he rejoined the AMTSO board for another two-year run. He also is the Vice Chair of the Executive Committee of IEEE ICSG. In 2018, Zwienenberg joined the Europol European Cyber Crime Center (EC3) Advisory Group as an ESET representative.
Zwienenberg has been a member of CARO since late 1991. He is a frequent speaker at conferences – among these Virus Bulletin, EICAR, AVAR, FIRST, APWG, RSA, InfoSec, SANS, CFET, ISOI, SANS Security Summits, IP Expo, Government Symposia, SCADA seminars, etc. – and general security seminars. His interests are not limited to malicious code but have broadened to include general cybersecurity issues and encryption technologies over the past years.
Mr. Eddy Willems
Eddy Willems is a worldwide known cyber security expert from Belgium. He is a board member of 3 security industry organizations, EICAR, AVAR and LSEC, and is the resident Security Evangelist at G DATA Cyberdefense.
He became a founding member of EICAR in 1991, one of the world’s first security IT organizations. Over the years he has served in many extra roles in different security industry organizations. Several CERTs, press agencies, print and online publications and broadcasting media, for example CNN, use his advice regularly. In October of 2013, he published his first book in Belgium and the Netherlands, entitled ‘Cybergevaar’ (Lannoo). A German translation followed afterwards and an English translation and update, Cyberdanger (Springer), was published in 2019. He is also co-author of the Dutch SF cyberthriller ‘Het Virus’ published in 2020. Eddy is a known inspiring speaker and is giving lectures and presentations (including TEDx) worldwide for a very diverse audience from children to experts.