Cyber Markets Round-up: Week 26, June '23: The good and bad: AI's disruptive power in cybersecurity

WHAT’S NEW

Are GPT-Based Models the Right Fit for AI-Powered Cybersecurity?

A growing number of cybersecurity vendors are integrating large-language model-based (LLM) tools into their offerings. Many are opting to use OpenAI’s GPT model.

Microsoft launched its GPT-4-powered Security Copilot in March and in April Recorded Future added a new research feature using OpenAI’s model trained on 40,000 threat intelligence data points.

LLMs are not the first type of AI that’s been integrated into cybersecurity products, with many Infosecurity Europe exhibitors – the likes of BlackBerry Cyber Security’s Cylance AI, Darktrace, Ironscales, and Egress – leveraging AI in their products.

However, although it’s difficult to say what AI algorithms cybersecurity vendors have used, they are very likely deterministic.

LLMs have been built on massive training sets. They’re also designed to guess the most probable words following a given prompt. These two features make them probabilistic and not deterministic, meaning they provide the most probable answer, not necessarily the right one. [InfoSecurity Magazine ]

FBI says artificial intelligence being used for 'sextortion' and harassment

The Federal Bureau of Investigation has warned Americans that criminals are increasingly using artificial intelligence to create sexually explicit images to intimidate and extort victims.

The bureau said the images appeared "true-to-life" and that, in some cases, children had been targeted.

The manipulation of innocent pictures to make sexually explicit images is almost as old as photography itself, but the release of open-source AI tools has made the process easier than ever. The results are often indistinguishable from real-life photographs, and several websites and social media channels that specialize in the creation and exchange of AI-enabled sexual imagery have sprung up in recent years. [Reuters ]

AI And Cyber Defense 2025: Decoding Defense Strategies

Precision Approach

• Employing an automated AI-powered threat intelligence platform that ingests real-time external signatures, tactics, techniques, and procedures (TTP). This platform works to be significantly faster in identifying and neutralizing phishing, malware, and other endpoint threats, evolving by learning from the methods of attacks.
• Companies need to implement continuous automated sensitive asset monitoring and alerting, from API inventory used across the company to personally identifiable information (PII) scanning to detect specific instances of plaintext exposure and alerting teams.
• Conducting AI-powered continuous code reviews, scanning for code exceptions, cross-site scripting flaws, code injection, buffer overflow, and more, and automatically replacing it with secure code while maintaining the functional integrity of the code.
• Leaders must look at this on a bigger picture and engage AI to detect malicious AI. Indirect prompt injection attacks highlight emerging threats where adversaries try to infiltrate Large Language Models (LLMs) indirectly, through AI being used to detect malicious polymorphic malware and more. [Forbes ]

Energy producer Suncor admits ‘cyber security incident’

A major Canadian energy producer and owner of the Petro-Can gas station network has acknowledged suffering an undefined cyber attack.

In a brief news release late Sunday, Suncor Energy said it “has experienced a cyber security incident. The company is taking measures and working with third-party experts to investigate and resolve the situation and has notified the appropriate authorities.

The attack on Suncor also comes after a cyber attack in April on an unnamed Canadian energy company. In that incident, stolen documents from the U.S. Pentagon quoted a pro-Russian group called Zarya claiming it compromised the operational technology side of a pipeline company. [IT World Canada ]

If you enjoyed this edition of the Cyber Markets Roundup, we are inviting you to connect with us on?Susbtack ?for free and let us build a community of like-minded individuals who are interested in cybersecurity, stocks, and financial markets.