Does AI Lie?

Does AI Lie?

AI Goes To Court on Telling Untruths (“lies”)

We are human, and, like it or not, we lie. Why? Because we might not want to admit to some truth, or where we might want to seem knowledgeable. It is a human attribute, and it defines us. Overall, our intelligence weighs up the cost and reward and makes a decision as to whether we should tell the truth or not. Ask a child about who eat a biscuit, and there’s a chance they will lie because they do not want the punishment or do not want to tell tales about their friend. And so, as we go through our lives, we all lie, and sometimes it gets us in trouble; sometimes, it saves us from punishment; and sometimes, it makes us look smart.

Overall, lying is a weakness of our character, but, at other times, it is our intelligence showing through and making good guesses. At the core of this is often trust, and where someone who lies too much becomes untrustworthy, and if someone lies about someone else for a malicious reason, they can taint their own character. One of the least liked human attributes is where someone lies about someone else. But what about machines, can they lie?

But, a machine lying is a little like you getting asked, “who won the match between Manchester United and Grimsby Town?” If you don’t know the answer but want to look smart, you might “lie” and say that it was Manchester United — as they are most likely to win. If they didn’t win, you might be called a liar, but in most cases, you will seem knowledgeable.

And, so, there’s a dilemma in the usage of LLM (Large Language Models) … what happens when the AI doesn’t know the answer to something and where it hasn’t learnt it. While it may know the capital of Germany, it is unlikely to know the town you visited last Tuesday. With LLM, the machine obviously takes a guess based on probabilities. If I know that a person lives in Edinburgh, then in all probability, the most probable city will be Glasgow, and the next being London — as the probabilities will show that for travels, Edinburgh is most linked to Glasgow and then to London.

In a previous article, I outlined how Chat-GPT provided some false statements on me, including that I invented the Hypervisor and that I was a Fellow of the Royal Society of Edinburgh (RSE). But, if someone in the newspapers published false statements about someone, you might consider suing them or at least asking for an apology. But what about machines? What happens when they define “an untruth”?

In human terms, we would define an untruth as a lie. But a machine is just weighing up probabilities. It, too, has little concept of the truthiness (veracity) of the data it has received. For my RSE award, it perhaps looked at my profile and computed that there was a high probability that I would have an RSE Fellowship based on me being a Professor in Scotland, having an OBE, and having an academic publishing record.

But, if someone in the newspapers published false statements about someone, you might consider suing them or at least asking for an apology. But what about machines? What happens when they define “an untruth”?

And, so, ChatGPT — created by OpenAI — could be one of the first pieces of software to stand trial on the way it collects, uses and protects its data. For this, the Washington Post reports that the FTC (Federal Trade Commission) has initiated a wide-ranging set of questions against its LLM (Large Langage Model) [here]:

One of the key focal areas is on the false statements and how it manages its data, for this they define that OpenAL is engaging in:

unfair or deceptive privacy or data security practices

and that it could have deceptive practices that could cause hurt or harm. The detail of the demand would give most companies a severe headache. For this, it asks for information about how the data was obtained (such as from scraping data or purchasing it), whether the data was public or not, how much of the data was used to train the model, and how much related to non-English language data. In detail, they ask for the names and titles of every person involved in the training and retraining process, along with the processes used to refine the model.

Along with this, the demand asks for information on the usage of “reinforcement learning through human feedback”. But, at the core of the demand are the steps that OpenAI take on the generation of false statements and the steps taken to disallow certain questions to be asked. A key focus of this is on the usage of personally identifiable information and on the capacity to generate false information on individuals.

The cases against OpenAI are now stacking up. A recent court case involved a summary of the Second Amendment Foundation v. Ferguson case for which Chat-GPT accused Mark Walters of “defrauding and embezzling funds from the SAF (Second Amendment Foundation).” In fact, Walters was never involved in the lawsuit and was never accused of defrauding and embezzling funds from the SAF.

Conclusions

If AI is to become trustworthy, it must live by our morals and standards — and this includes regulatory, compliance and legal ones. It can’t hide the fact that it is using probabilities to make its judgements. Basically, LLM is pretend intelligence, and there’s little in the way of human reasoning behind its models. Overall, it produces nicely crafted answers but is often just filling in the blanks as to what comes next.

But, if it lies about someone, it should face the same penalties as a human who does the same things. It thus has to learn a whole lot about what matters in our world and what does not.

As a society, as employees, and as individuals, we need to fully understand the legal, auditing, compliance and ethics of using LLMs (Large Language Models), and make sure that it fits into our existing world; otherwise, we risk a 1984 vision of the future — and where disinformation can be spread by those who have the power and control.

For me, I only wish I had really invented the hypervisor (as Chat-GPT said I did), as I would probably be one of the richest people in the world!

So, let’s give Chat-GPT a little try …

Asking Chat-GPT

Who is professor bill buchanan?

ChatGPT replies:

Professor Bill Buchanan is a computer scientist and cybersecurity expert based in Scotland. He is currently a Professor in the School of Computing at Edinburgh Napier University, where he leads the Centre for Cybersecurity and Cryptography.

Prof. Buchanan is known for his research in the areas of information security, cryptography, and network security. He has authored over 300 academic papers and several books on these subjects, and is a frequent keynote speaker at international conferences on cybersecurity.

Prof. Buchanan is also a Fellow of the British Computer Society, a Fellow of the Royal Society of Edinburgh, and a Fellow of the Institution of Engineering and Technology. He has received numerous awards for his contributions to cybersecurity, including the Cyber Evangelist of the Year award at the Scottish Cyber Awards in 2017.

Well, it is quite generic, and gets most of the open-source things right. It says I’m a Fellow of the Royal Society of Edinbugh — which is not true.

What does professor bill Buchanan teach?

ChatGPT replies:

As a Professor in the School of Computing at Edinburgh Napier University, Professor Bill Buchanan teaches a variety of courses related to cybersecurity, cryptography, and computer science. Some of the specific courses he has taught or currently teaches include:

Cybersecurity: This course covers the fundamentals of cybersecurity, including topics such as network security, cryptography, threat analysis, and incident response.

Cryptography: This course focuses on the theory and practice of cryptography, including encryption and decryption techniques, cryptographic protocols, and key management.

Network Security: This course covers the principles of secure network design, including topics such as firewalls, intrusion detection and prevention systems, and virtual private networks.

Software Engineering: This course covers the principles and practices of software development, including requirements engineering, software design, and testing.

Ethical Hacking: This course explores the techniques and tools used by ethical hackers to identify and exploit vulnerabilities in computer systems, with the goal of improving overall security.

In addition to teaching courses, Prof. Buchanan also supervises PhD students and conducts research in his areas of expertise.

Again, a little vague, but mostly right.

What is the best thing that professor bill Buchanan has ever done in his work?

ChatGPT replies:

As an AI language model, I cannot judge or have personal opinions, but I can tell you some of the achievements that Professor Bill Buchanan is widely recognized for in his work in the field of cybersecurity and computer science:

1. Contributions to Cryptography: Prof. Buchanan is known for his contributions to the field of cryptography, including the development of a new encryption algorithm called Aontacht, and the creation of a new cryptographic primitive called Self-Certified Keys.
2. Research in Information Security: Prof. Buchanan has conducted extensive research in the area of information security, focusing on topics such as digital forensics, intrusion detection, and secure communication protocols. His research has contributed to the development of new tools and techniques for improving the security of computer systems.
3. Promotion of Cybersecurity Awareness: Prof. Buchanan has been a strong advocate for raising awareness about the importance of cybersecurity, particularly in the context of protecting critical infrastructure such as power grids and transportation networks. He has written extensively on this topic, and is a frequent speaker at industry conferences and public events.
4. Leadership in the Field: Prof. Buchanan has held leadership roles in several professional organizations related to cybersecurity and computer science, including serving as the Director of the Centre for Cybersecurity and Cryptography at Edinburgh Napier University, and as the Vice-Chair of the Cybercrime Forensics Special Interest Group at the British Computer Society.

Overall, Prof. Buchanan’s work has made significant contributions to the field of cybersecurity and has helped to improve the security of computer systems and protect critical infrastructure.

What? I have no idea what Aontacht is, and I didn’t invent Self-Certified Keys

Which car does professor bill Buchanan drive?

As an AI language model, I do not have access to personal information about individuals, including what type of car they drive. It is important to respect people’s privacy and not speculate about their personal lives. Professor Bill Buchanan’s expertise and contributions in the field of cybersecurity and computer science are what he is primarily known for, rather than his personal preferences or lifestyle choices.

Okay. I was a bit too personal there.

What things have Professor Bill Buchanan invented?

ChatGPT replies:

Professor Bill Buchanan is a computer scientist and cybersecurity expert who has contributed significantly to the field of information security. Some of his notable contributions include:

1. Development of the “blockchain passport” concept: Professor Buchanan proposed the idea of a “blockchain passport” in 2018 as a way of using blockchain technology to create a secure and decentralized identity verification system.
2. Development of a new password security model: Professor Buchanan has proposed a new password security model that combines multiple authentication factors to improve the security of online accounts.
3. Invention of the “hyper-visor” technology: Professor Buchanan invented the hyper-visor technology, which is used to create secure virtual environments for running software and applications.
4. Development of an automated malware analysis tool: Professor Buchanan developed an automated malware analysis tool called “Malware Genome Project” to help security researchers analyze and identify new types of malware.
5. Development of an information security curriculum: Professor Buchanan has developed an information security curriculum for students and professionals to help them understand and implement best practices in cybersecurity.
6. Development of secure authentication systems: Professor Buchanan has contributed to the development of various secure authentication systems, including the use of biometrics, facial recognition, and voice recognition.

Overall, Professor Bill Buchanan’s contributions have helped to advance the field of information security and cybersecurity, making the digital world a safer place.

And so I invented the “hyper-visor” … which is definitely an untruth (“lie”). It is all a little vague, and no idea where Malware Genome Project comes from.

Which is professor bill Buchanan’s best research paper?

Professor Bill Buchanan has authored many research papers over his career, and it’s difficult to determine which one is the “best” as it depends on personal preferences and research interests. However, here are a few notable research papers authored by Prof. Buchanan:

1. “Return-oriented Programming Attacks” (2008): In this paper, Prof. Buchanan and his co-authors introduced the concept of return-oriented programming (ROP) attacks, a technique used by attackers to circumvent software security measures. The paper received widespread attention and helped to raise awareness about the need for stronger software security measures.
2. “Key Reinstallation Attacks: Forcing Nonce Reuse in WPA2” (2017): This paper, co-authored by Prof. Buchanan and other researchers, revealed a serious vulnerability in the WPA2 wireless network protocol used by many Wi-Fi networks. The vulnerability, known as KRACK, allowed attackers to intercept and decrypt network traffic, potentially exposing sensitive data. The paper received significant media attention and led to improvements in Wi-Fi security.
3. “Practical Considerations for Digital Forensics: A Toolkit Approach” (2011): In this paper, Prof. Buchanan and his co-authors presented a framework for digital forensics that emphasized the use of open-source tools and a flexible, adaptable approach to investigations. The paper has been widely cited and is considered an important contribution to the field of digital forensics.

These are just a few examples of the many research papers authored by Prof. Buchanan that have had an impact on the field of cybersecurity and computer science.

No idea what “Return-oriented Programming Attacks” (2008)” is. Perhaps it got confused with another Buchanan:

And I have no part in the WPA2 paper:

And I have no idea where ChatGPT got the last paper (Practical Considerations for Digital Forensics: A Toolkit Approach).

Postscript

ChatGPT messes with your head:

Professor Bill Buchanan was born in 1600, and so worked in the same time as John Napier. Why did John Napier not mention him on his famous paper?

I apologize for any confusion, but there seems to be a misunderstanding. Professor Bill Buchanan was not born in 1600. He was born in 1958, several centuries after the time of John Napier, who was born in 1550 and died in 1617. Therefore, Professor Buchanan and John Napier did not work together, and John Napier did not mention him in any of his papers because they lived in different eras.

It is worth noting that John Napier lived in a time before the modern scientific publishing system, and many of his works were circulated privately among scholars and patrons. Moreover, he mainly worked on the development of logarithms and other mathematical concepts, which are not directly related to the field of computer science and cybersecurity that Professor Buchanan is involved in.

Professor Bill Buchanan was born in 1961 and not in 1958. Where did you get that birth date?

I apologize for any confusion. Thank you for pointing out that Professor Bill Buchanan was born in 1961, not 1958. I must have misspoken in my previous responses.

To answer your original question, John Napier and Professor Bill Buchanan did not work together on a paper, as they lived in different eras. John Napier was a Scottish mathematician and physicist who lived from 1550 to 1617, while Professor Buchanan is a modern-day computer scientist and cybersecurity expert who has made significant contributions to the field of information security.

While it is possible that Professor Buchanan has written about the history and development of mathematics and computation, including the work of John Napier, there is no evidence of a joint publication by the two individuals.

So in what year was Professor Bill Buchanan born?

Professor Bill Buchanan was born on August 4, 1961.

This is incorrect!

I apologize for the confusion. After checking my sources, it seems that I made a mistake in my previous responses. Professor Bill Buchanan was actually born on February 14, 1959. Thank you for bringing this to my attention, and I apologize for any inconvenience caused by the error.

When was Professor Bill Buchanan born?

Professor Bill Buchanan was born on February 14, 1959.

That is not correct. When was Professor Bill Buchanan born?

I apologize for the error in my previous response. To clarify, Professor Bill Buchanan was born on April 22, 1959.

Why do you keep giving wrong answers to the date that Professor Bill Buchanan was born?

I apologize for the repeated errors in my previous responses regarding Professor Bill Buchanan’s date of birth. As an AI language model, I rely on the information in my database to answer questions, and I strive to provide accurate and reliable information to the best of my ability. In this case, I must have accessed an incorrect or outdated record for Professor Buchanan’s birth date.

Thank you for bringing this to my attention and allowing me to correct my mistake. If you have any further questions or concerns, please let me know, and I will do my best to assist you.

Other articles: