article thumbnail

When AIs Start Hacking

Schneier on Security

As I lay out in a report I just published , artificial intelligence will eventually find vulnerabilities in all sorts of social, economic, and political systems, and then exploit them at unprecedented speed, scale, and scope. These are artificial and constrained, with all of the rules specified to the AI. Not for long.

Hacking 357
article thumbnail

Why BYOD Is the Favored Ransomware Backdoor

eSecurity Planet

Secure web gateways (SWG) can sometimes inspect traffic and potentially detect increased and anomalous traffic. Anomaly detection, often enhanced using artificial intelligence, can both improve detection and block activity, but only when traffic routes through these devices.

Insiders

Sign Up for our Newsletter

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

article thumbnail

Vulnerability Recap 4/1/24: Cisco, Fortinet & Windows Server Updates

eSecurity Planet

Vendors and researchers disclosed a wide range of vulnerabilities this week from common Cisco IOS, Fortinet, and Windows Server issues to more focused flaws affecting developers (PyPI), artificial intelligence (Ray, NVIDIA), and industrial controls (Rockwell Automation).

article thumbnail

How Steganography Allows Attackers to Evade Detection

eSecurity Planet

While steganography is often considered something of a joke in capture-the-flag (CTF) events and other cybersecurity defense activities, it can happen in real attacks and can take security defenses by surprise simply by using another layer of cover. Steganography Techniques Evolve Quickly.

article thumbnail

Defining Good: A Strategic Approach to API Risk Reduction

Security Boulevard

Organizations first looked to augment their existing web application security tools and processes to “address” API security. Unfortunately, the security challenges associated with APIs can't be solved by simply updating existing testing tools and edge security defenses to check-the-box technologies that claim to provide "API security."

Risk 57
article thumbnail

2019 Recap: A Year to Remember

McAfee

During this year’s RSA conference, VP and Chief Technology Officer Steve Grobman and Chief Data Scientist Dr. Celeste Fralick took the main stage to discuss how the industry needs to think about artificial intelligence, its power, and the possible ways it can be used against us.

article thumbnail

Four ways to stay ahead of the AI fraud curve

SC Magazine

While AI increasingly gets used to automate repetitive tasks, improve security and identify vulnerabilities, hackers will in turn build their own ML tools to target these processes. Secure and manage AI to prevent malfunctions.