Explore the intersection of AI and cybersecurity. Stay informed on AI-driven security trends, tools, and threats in the ever-evolving digital landscape.
Search across headline titles and summaries.
Background for this topic.
Artificial intelligence (AI) describes computer systems that perform tasks such as recognizing patterns, making predictions, understanding language, or generating content. In security reporting, the term commonly includes machine-learning models used for detection and analysis, as well as generative AI applications that produce text, code, images, or other outputs.
AI can help analyze security telemetry, prioritize vulnerabilities, and support investigations, but its outputs can be wrong or manipulated. Important attack surfaces include prompt injection that steers an application into unintended actions, sensitive data being exposed through prompts or model outputs, and excessive permissions granted to AI systems that use external tools. Models can also be degraded by poisoned training data or evaded with carefully crafted inputs. Practitioners should protect training and operational data, limit model access and tool permissions, test for adversarial behavior, and require appropriate human validation before high-impact decisions.
Also: Addressing Scotland's Cybercrime Surge; NOC and SOC ConvergenceIn the latest weekly update, ISMG editors discussed the convergence of the NOC and SOC functions, Scottish Police efforts to address the escalating challenge of cybercrime in Scotland, and why OpenAI is pushing to dismiss certain aspects of The New York Times lawsuit.
Dark Reading's roundup of strategic cyber-operations insights for chief information security officers and security leaders. Also this week: SEC enforcement actions, biometrics regulation, and painful encryption changes in the pike.
Lack of AI model visibility and security puts the software supply chain security problem on steroids.
A growing thicket of privacy laws regulating biometrics is aimed at protecting consumers amid increasing cloud breaches and AI-created deepfakes. But for businesses that handle biometric data, staying compliant is easier said than done.
How zero trust controls and Google AI can strengthen your organization’s defences Webinar Dealing with cyber security incidents is an expensive business. Each data breach costs an estimated $4.35 million on average and it's not as if the volume of cyber attacks is falling - last year, they rose by 38 percent according to Google Cloud.…
Convincing phishing emails, synthetic identities, and deepfakes all have been spotted in cyberattacks on the continent.
Agent Robert K. Tripp on FBI's Approach to Deepfakes, Nation-State Election ThreatsThe U.S. presidential election is still eight months away, but the FBI is already seeing its share of cyberattacks, nation-state threats and AI-generated deepfakes. According to FBI Agent Robert K. Tripp, "We're no longer considering threats as a what-if situation; it's happening now."
Brave Software is the next company to jump into AI, announcing a new privacy-preserving AI assistant called "Leo" is rolling out on the Android version of its browser through the latest release, version 1.63. [...]
GitHub has enabled push protection by default for all public repositories to prevent accidental exposure of secrets such as access tokens and API keys when pushing new code. [...]
Microsoft has released the Windows 11 'Moment 5' update for versions 23H2 and 22H2, starting the rollout of new features, such as Windows Copilot skills and plugins, Voice Access, AI enhancements for ClipChamp and Photos, and Narrator improvements. [...]
Infoblox said Savvy Seahorse uses fake ChatGPT and WhatsApp bots to lure victims
The finding underscores the growing risk of weaponizing publicly available AI models and the need for better security to combat the looming threat.
AI service Cutout.Pro has suffered a data breach exposing the personal information of 20 million members, including email addresses, hashed and salted passwords, IP addresses, and names. [...]
Talk about gone in 60 seconds Computer scientists at the University of Maryland have developed an efficient way to craft prompts that elicit harmful responses from large language models (LLMs).…
Scientists reveal automated adversarial prompt generation too powered Nvidia visualization accelerator Computer scientists from the University of Maryland have developed an efficient way to generate adversarial attack phrases that elicit harmful responses from large language models (LLMs).…
At least 100 instances of malicious AI ML models were found on the Hugging Face platform, some of which can execute code on the victim's machine, giving attackers a persistent backdoor. [...]
Lawsuit alleges it misled investors with claims new AI products were 'facilitating greater platformization' and more Palo Alto Networks (PAN) is facing a proposed class action lawsuit that alleges investors were deceived about the traction of its platform tactics and hurt by an unexpectedly low billings forecast that crashed the share price.…
It's time for companies to look at what they're processing, what types of risk they have, and how they plan to mitigate that risk.
UK Information Commissioner John Edwards explains how the ICO is working to provide clarity around the lawful use of AI