“Collective wisdom” needs to keep up as hackers harness AI

The World Economic Forum (WEF) urges companies to implement novel cybersecurity measures at the speed of AI implementation.

“AI introduces risk at the same rate as it introduces efficiencies,” WEF said, highlighting the exposure of intellectual property to generative AI engines like ChatGPT.

WEF also referred to a report by the security firm Check Point, saying that attackers leverage AI to identify and exploit flaws at scale.

Namely, the report discusses HexStrike AI – an AI brain that orchestrates over 150 specialized AI agents to autonomously scan, exploit, and persist within targets. Attackers even attempted to use HexStrike-AI to go after recent zero-day flaws.

“These vulnerabilities are complex and require advanced skills to exploit. With Hextrike-AI, threat actors claim to reduce the exploitation time from days to under 10 minutes,” Check Point said.

As it often is with cybersecurity tools designed to aid the good guys, Hexstrike-AI was supposed to be a defender-oriented framework for red teams and security researchers.

However, immediately after its release, malicious hackers started looking for ways to weaponize it.

“This marks a pivotal moment: a tool designed to strengthen defenses has been claimed to be rapidly repurposed into an engine for exploitation, crystallizing earlier concepts into a widely available platform driving real-world attacks,” Check Point said.

AI applications, as WEF aptly noted, are also used to craft more realistic and appealing phishing lures to extract people’s credentials and other sensitive information.

Given that one of the AI’s key features is its adaptability, security measures should also be able to change over time.

“Flexibility is essential. An “open garden” approach to security enables organizations to build and customize stacks that align with their unique needs, while ensuring that technologies work together effectively. Security solutions should be context-aware, future-ready and able to integrate across hybrid environments,” WEF said.

Today, AI can be used to faster detect threats, automate remediation, and “compress an approval chain.”

“Localized breaches can turn into cascading failure. AI introduces risk at the same rate as efficiency. If not planned for and managed in a thoughtful manner, these risks can impact your brand in profound ways,” WEF said, not giving us much to cling on to what organizations should do to mitigate threats.

“Collective wisdom” needs to keep up

Meanwhile, researchers, business leaders, educators, and students recently gathered for the inaugural MIT Generative AI Impact Consortium (MGAIC) Symposium to attempt to answer a bigger question: What does the future hold for generative AI?

Yann LeCun, chief AI scientist at Meta, believes that significant advances in AI will most likely come through “world models” that learn similarly to how kids learn – by interacting with the world around them.

“A 4-year-old has seen as much data through vision as the largest LLM. … The world model is going to become the key component of future AI systems,” he said.

Scientists and engineers were urged to design guardrails to keep future AI systems on track.

“This is a pivotal moment — generative AI is moving fast. It is our job to make sure that, as the technology keeps advancing, our collective wisdom keeps pace,” said Anantha Chandrakasan, MIT's Provost and the Vannevar Bush Professor of Electrical Engineering and Computer Science.

Unlock more exclusive Cybernews content on YouTube