Hacking with ChatGPT: Five A.I. Based Attacks for Offensive Security

ChatGPT may represent one of the biggest disruptions in modern history with it's powerful A.I based chatbot. But within weeks of ChatGPT's release, security researchers discovered several cases of people using ChatGPT for everything from malware development to exploit coding. In this video, take a look at the five ways attackers are utilizing ChatGPT for wrong doing.
0:14 Intro to ChatGPT / Natural Language Processing (NLP) & GPT
1:28 Using ChatGPT for Vulnerability Discovery
1:56 Vulnerability Prompts to Utilize
3:10 Writing Exploits
3:35 Exploit Prompts to Utilize
4:33 Malware Development
5:00 Malware Examples (Stealers, Command & Control)
5:42 Polymorphic Malware Development Using ChatGPT
6:21 A.I. Based Phishing using NLP (Natural Language Processing)
7:20 ChatGPT Advantages over Traditional Phishing Messages
7:41 Custom Messages Using GPT-3
8:04 Using Macros and LOLBINs
9:33 GPT-3 vs GPT-4 (Coming Soon)
9:56 Cybersecurity Considerations and Predictions

[@meliois3910]

And with GPT-4 coming, this is going to be a race for A.I mastery between offsec and defensive sec

[@fokyewtoob8835]

Been saying this. Gonna be like a constant cat and mouse between both groups and their AI. I’m excited

[@MisterK-YT]

Well this didn’t turn out to be true lol. ChatGPT is becoming wildly more restrictive every day.

[@fokyewtoob8835]

@@MisterK-YT “this didn’t come true” has been literally 4 months lol relax

[@Wejustwannatravel]

This is literally the perfect plug for Vectra AI ✅

[@riverroad935]

Thank you very much for the nice and easy-to-understand short lesson.

[@Hacktheplanet_]

Interesting videos

[@jonnysokkoatduckdotcom]

What happens when version 5 comes out soon this year 2023 a full new version

[@djhakdabeat6650]

Awesome presentation!

[@leonardobq1982]

Great project explaining security hot topics! Congrats!

[@brothers931ify]

Do you have a video specifically About how To stop and to identify someone who is hacking you? How to prove you are being target hacked?

[@amparoconsuelo9451]

ChatGPT & Hacker VS. ChatGPT & Cybersecurity

[@biocode2856]

Yeah use a PDF write yourself in as ministrator or co

[@montagistreel]

*GPT-3

[@cybersamurai99]

superb friend!

[@user-vp5ji7qv6k]

I love this Channel, Im such a nerd.

[@MrNightowl1980]

Thank you so much for posting these videos. Your way of explaining is easy to digest!

[@TrusteestDesiChhora]

Very helpful video thank you 😊❤

[@kristaylor3175]

This was a great video! Thanks

[@pitercompany7344]

Can you explain it one by one and much deeper !

[@ruhruhruhruhruheisjsij]

If script kiddies do use this though , most of them probably don't realize that basically all of their chats are saved and backed up on the openAI servers lol and there is no way to remove it. Leaves a massive digital foorprint

[@FlutterDev1337]

There's tgpt. No PII/API needed.

[@-UE-PR0]

@@FlutterDev1337what you mean

[@kadhamba_topic6941]

vpn's plus fake email's would help

[@-UE-PR0]

I’m confused. Just because T-GPT doesn’t use any api, it still runs in openai cloud server. So if and when openai shuts down you can no longer use tgpt

[@phabeondominguez5971]

13'th

[@montagistreel]

artificial neurons, not atoms

[@fdhin]

100x more powerful? Painful mistake in such otherwise well produced content

[@TheCISOPerspective]

This number has been debated, which is why I said "reported". The source came from several online articles:
-finance.yahoo.com/news/whats-under-hood-microsofts-bing-193336328.html?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAJJt4FmxcVnxf_BpPZKHn-QBTHtAGAwLvD6No76jLUNnTkpKCB_mETMd4y8sveMTV7mIU3mI8mGh1l5hMFA8k4IdcB5x8HHFR3HedZCLFoj9vx6ACuXneYXGlYQzz-uFF6NEk0J2yzEQeOTOt0N6qlV-4yYTZPDog-Ww-KfPv6RF
-medium.com/geekculture/gpt-4-100x-more-powerful-than-gpt-3-38c57f51e4e3
-indianexpress.com/article/technology/tech-news-technology/chatgpt-4-release-features-specifications-parameters-8344149/

[@marccornelius6989]

*promosm* ✅