Latest cybersecurity reporting from selected sources.

Threat Actor Silently Forwarded Sensitive Emails Matching Strategic TopicsGoogle says Chinese espionage group UNC6508 compromised REDCap environments at North American research institutions, deployed custom malware, stole credentials and covertly forwarded strategically relevant emails through abused compliance rules to support long-term intelligence collection.

Mandiant: 68% of Targets Were Higher Ed Institutions Running PeopleSoftShinyHunters exploited a critical zero-day in Oracle PeopleSoft to breach more than 100 organizations globally, researchers at Mandiant and Google's Threat Intelligence Group said, with universities and colleges accounting for the majority of confirmed targets in the active extortion campaign.

Google Says Criminals Used AI to Discover and Code ExploitA cybercriminal group came close to launching a mass attack earlier this year, armed with a software exploit that an AI model had built from scratch, said Google researchers. Google said it worked with the affected vendor to patch the flaw before an attack could be launched.

Researchers Say PromptSpy Automates Persistence on Infected DevicesA newly discovered Android malware strain, "PromptSpy," is using Google's Gemini generative artificial intelligence model to automate part of its persistence mechanism, marking what researchers describe as the second known case of AI-driven mobile malware.

Signs Point to Multiple Exploit Chains, One Including a Zero-Day, Being EmployedData-stealing attacks targeting Oracle E-Business Suite, for which an affiliate of Russian-speaking Clop ransomware group is claiming credit, appear to have begun by August and involved multiple attack chains, of which one targeted a zero-day vulnerability, report Google threat researchers.

'More Research' Needed, Says GoogleGoogle DeepMind expanded its risk framework to cover scenarios where artificial intelligence models might manipulate people or resist shutdown, marking the company's most explicit warning yet about potential misalignment. DeepMind frames many of these scenarios as a malfunction.

'Widespread Data Theft Campaign' Compromised Many Drift OAuth Tokens, Warn ExpertsThreat researchers report that "a widespread data theft campaign" traces to attackers stealing OAuth access tokens for applications integrated with Salesloft's AI chatbot Drift, then exfiltrating data. Victims include Salesforce customer Zscaler. Google Workspace instances were also breached.

Experts Aim to Probe How AI Models Reason, and Why It MattersAI researchers from OpenAI, Google DeepMind and Anthropic and others have urged deeper study into chain-of-thought monitoring, a technique to track how reasoning models arrive at answers. Their joint paper warns that transparency may erode if not prioritized.

Attackers Can Trick Gemini Into Displaying Deceptive Messages, Researchers WarnAttackers can hide malicious instructions inside emails to trick Google's Gemini into delivering falsified summaries with deceptive messages to end users, researchers warn. Google said it's continuing to put multiple defenses in place to combat these types of prompt injection attacks.

Google’s Mandiant Warns About Remote Attacks Disrupting Grid StabilityVulnerabilities in networked devices programmed to instantaneously trip power grid substation circuit breakers could be the means hackers use to cause the next blackout, warn researchers. There are "systemic patterns across substations, utilities and industrial sites worldwide," Mandiant warned.

Defunct Ransomware Group's Diaspora Includes Hackers With Focus on Microsoft TeamsBased on intelligence gleaned from the leak of Black Basta ransomware messages, researchers are warning organizations to beware phishing attacks launched via Microsoft partner domains and via Teams, as well as the targeting of personal Google accounts accessed via corporate devices.

Malware Targets Western Officials, NGOs and JournalistsRussian cyber espionage hackers are using a new malware strain dubbed "Lostkeys" in a targeted espionage campaign aimed at Western officials, NGOs and journalists. Google researchers attribute Lostkeys to the threat group Coldriver, an operational unit within the Federal Security Service.

Attacker With Project Access Could Have Retrieved Private Images, Researchers SaidGoogle has fixed Google Cloud Platform vulnerability attackers could have exploited to gain unauthorized access to private container images, due to access restriction shortcomings. Researchers said the flaw highlights how services built atop other services can pose unexpected security risks.

Google Expects Tactics to Spread; Global Targets and Other Services at RiskRussian nation-state hackers are using phishing attacks to target Ukrainian users of the chat app Signal, say security researchers. Rather than circumventing Signal's end-to-end encryption via a cryptographic attack, attackers use malicious prompting to prod victims into exposing messages.

Russia, China, Iran and North Korea Tapping Cybercrime Services, Google SaysThe cybercrime-as-a-service economy continues to power ransomware and other criminal enterprises, as well as serve as "an accelerant for state-sponsored hacking," collectively posing an increasing risk to Western national security, cybersecurity researchers warn.

Summit to Focus on Open-Source, AI Governance and DevelopmentThe historic presidential Élysée Palace in Central Paris will host world leaders, tech CEOs and researchers for the French AI Action Summit, a two-day event that will commence on Monday. U.S. Vice President JD Vance, OpenAI CEO Sam Altman and Google's Sundar Pichai will be on hand.

Developers Listed as Public Contact Points Targeted in Phishing CampaignA supply chain attack that subverted legitimate Google Chrome browser extensions to inject data-stealing malware is more widespread than security researchers first suspected. So far researchers have identified 36 subverted extensions collectively used by 2.6 million people.

Multimodal Agentic AI Delivers Speed, Tools, and Research PrototypesGoogle's latest AI model can natively process and output text, images and audio in the search giant's push toward more autonomous reasoning, planning and action. The company said Gemini 2.0 is designed for applications ranging from development and gaming to research and everyday assistance.

Hackers can potentially use AI to manipulate data that's generated and shared by some health apps, diminishing the data's accuracy and integrity, said Sina Yazdanmehr and Lucian Ciobotaru of cybersecurity firm Aplite, describing a recent research project involving Google Health Connect.

One Vulnerability Had Been Undiscovered for Two Decades, Researchers SaidGoogle researchers used an AI-powered fuzzing tool to identify 26 vulnerabilities in open-source code repositories, some of which had been lurking undiscovered for several decades. Each was found with AI, using AI-generated and enhanced fuzz targets, Google said.