Artificial Intelligence: Boom or Bust?  Part 2

Small robot whispering in a human's ear

By Guest Blogger, Chris Bonatti, Cybersecurity Consultant, IECA of Casper

Missed Part 1? Here’s the link https://cyberwyoming.org/artificial-intelligence-boom-or-bust-part-1/

AIs For Persuasion

It takes little imagination to understand how AIs could be leveraged for deception and social engineering. However, the practical possibilities range much further… to potentially include AI-enabled propaganda campaigns. AIs such as ChatGPT excel at producing written material, and we’ve also seen AIs such as Microsoft’s VALL-E system that can closely simulate any voice given a three-second audio sample. We’ve also seen AIs leveraged to create bogus photos and videos… so-called “deepfakes”. Such technology has already been leveraged to defeat voice recognition systems used for banking authentication. The negative possibilities for disinformation, manipulation, and social engineering are obvious. What is less obvious is that AIs could be leveraged to mislead, and literally control our societies via legislation. Well-known cryptographer and public interest technologist, Bruce Schneier, has co-authored several essays highlighting the risks of AI-enabled political lobbying. AIs could be capable of crafting lobbying strategies with compelling power. Legislation crafted by AIs could be equally compelling… and equally deceptive.

AIs Bolstering Defense

Not to present only doom and gloom, many industry watchers have predicted that AIs could prove to be a real benefit to overworked and overstressed cyber defenders. There have been some examples of this too. Researchers in Sweden recently used an AI to crack the NIST-recommended Post-Quantum Encryption (PQE) algorithm CRYSTALS-Kyber. Such a development could prevent the widespread deployment of flawed algorithms. An AI also managed to aid Abnormal Security (see ‘abnormalsecurity.com’) to detect and thwart an attempted Business Email Compromise (BEC) fraud that might have cost a real estate investment company over $36 million. AIs can analyze large amounts of data and identify patterns or anomalies much faster and with greater accuracy than humans. They can also detect and respond to security threats in real time, reducing the time it takes to identify and remediate security incidents. Consequently, many security companies are scrambling to incorporate AI features into their products. For example, Microsoft just introduced a version of its Security Copilot tool that leverages OpenAI’s GPT-4 engine.

To be continued in part 3 …

Share:

Register to Receive the Tech Joke of the Week!

This Week's Joke:

How many programmers does it take to change a light bulb?

None, it is a hardware problem!

More Posts: