February has been an eventful month in the tech sector as ransomware operator LockBit, which has accrued billions of dollars in ill-gotten gains in its approximately four-year history, was taken down by an international task force spearheaded by the UK’s National Crime Agency.
Elsewhere, ChatGPT suffered a major setback as users noticed the industry-leading chatbot had taken to answering prompts with complete gibberish. Though the issue was fixed within a day of being reported, it has raised important questions about the reliability of the service and the technology that underlies it.
In this episode, Jane and Rory welcome back Ross Kelly, ITPro’s news and analysis editor, to explore February’s big cyber security story and discuss what ChatGPT’s moment of madness means for generative AI.
Highlights
“It's not unprecedented. We have seen these sort of instances before, we've seen it with the Qakbot, Emotet takedowns. We've seen it with ALPHV as welI. I think what's really interesting here is that law enforcement agencies are becoming really aggressive and there's a lot of global coordination here. We've got the FBI, NCA, a lot of international partners coordinating on this.”
“The NCA is obviously going to emphasize the impact of this and a positive thing in terms of taking LockBit down. LockBit are going to really try and underplay as far making it seem like they're still a really serious threat for organizations.”
Discover the real-world insight about the products to watch in 2024 from our in-house experts
“There were a few instances where people were asking for coding advice and were, again, getting nonsensical answers back. If you're working to a deadline [it] might not be the best thing in the world when you're getting ChatGPT essentially just screaming garbage at you.”
Footnotes
- LockBit could be done and dusted after NCA operation gained access to admin environments, source code, and affiliate info
- LockBit rises from the ashes, but will it pack the same punch as before?
- Life after LockBit: A fragmented landscape and wayward affiliates will still cause chaos for enterprises
- Europol takes down 'dangerous' Emotet botnet
- Qakbot forced offline, but history suggests it probably won’t be forever
- Everything we know so far about the rumored ALPHV 'takedown'
- History tells us ALPHV will likely recover from recent takedown
- LockBit remains most dangerous ransomware despite fall in attacks
- The 'Big Three’ ransomware groups are losing their grip on the industry as gangs begin to fracture, study shows
- ChatGPT has gone haywire, and nobody knows why
- ChatGPT gives wrong answers to programming questions more than 50% of the time
- OpenAI's Sam Altman: Hallucinations are part of the “magic” of generative AI
- Google’s latest warning over generative AI shows that ChatGPT has it cornered
Subscribe
