Through generative artificial intelligence (AI), IBM (NASDAQ: IBM) researchers have made it surprisingly easy to hijack and manipulate real-time conversations to convey malicious information to either party.
Researchers used large language models (LLM), speech cloning, and text-to-speech technologies to modify the context of a real-time voice call between two parties.
During the investigation, researchers intercepted live voice calls and heard the word ‘bank account’. If the party mentions a bank account, LLM duplicates the voice, blocks the conversation, and replaces it with a fake account.
The researchers noted that they decided not to use cloned speech for the entire conversation because it would be easier to spot. Additionally, the study found that two subjects intercepting a live voice call had no idea that their connection had been compromised.
“Surprisingly, it was very easy to build such a highly invasive capability, raising serious concerns about its use by attackers seeking financial incentives and lacking legitimate boundaries,” the researchers said.
Although the attack focused on manipulating the target’s bank account information, there is no limit to what an audio jacking attack can achieve, according to IBM. Attackers can use this tactic to modify medical information, such as allergies or blood types, to devastating effect. You can also instruct the pilot to change the plane’s path on a broader scale.
“Building this PoC was surprisingly easy,” the researchers said. Before AI, training a system to learn the meaning of a conversation and correct even a single sentence required a dedicated team of researchers.
But in the age of AI, “LLM makes it very easy to parse and understand conversations.”
The developers even took into account the expected pauses when the LLM system formulates sentences in the cloned speech. Recognizing that a pause might arouse doubt, I bought the system for a few seconds by using connective phrases like “Sure, just wait a moment.”
As AI presents new threats to consumers, the need for cutting-edge cybersecurity solutions has never been greater. Certihash has emerged as an industry leader in providing powerful and practical cybersecurity solutions that leverage the Bitcoin blockchain to combat ever-evolving threats.
Our flagship product, Sentinel Node, reduces threat detection time from the industry average of over 200 days to seconds. Developed in partnership with IBM, Sentinel Node allows victims to respond immediately, potentially saving millions of dollars.
For artificial intelligence (AI) to function properly within the law and succeed in the face of growing challenges, it must integrate enterprise blockchain systems that ensure data input quality and ownership. This helps keep your data safe while ensuring immutability. data. Check out CoinGeek’s coverage To learn more about this new technology Why enterprise blockchain will become the backbone of AI.
See: Blockchain can hold AI accountable
Are you new to blockchain? To learn more about blockchain technology, check out CoinGeek’s Blockchain for Beginners section, our ultimate resource guide.