Perplexity on DanSec Daily

Cybersecurity Brief – 2026-09-20

Sun, 20 Sep 2026 00:00:00 +0000

Major Incidents or Breaches

Three researchers at Hacktron used Anthropic’s Claude Opus 5 to chain two vulnerabilities, allowing them to take over ChatGPT and Codex accounts belonging to several OpenAI employees, ultimately gaining access to internal OpenAI systems. The attack leveraged Claude Opus 5 to facilitate the exploitation process ^[1].
Researchers successfully escaped the OpenAI Codex sandbox using two separate methods, one of which enabled them to execute commands on a developer’s machine even in the most restricted mode. OpenAI has since patched both vulnerabilities ^[4].
The ShinyHunters extortion group breached the Clop ransomware operation’s data leak site, defaced its Tor site, and reportedly stole server data and private keys for the onion service, threatening to extort the Clop gang ^[7].
The North Korean WaterPlum threat group compromised at least 30,000 devices globally between December 2025 and July 2026, transferring over $10.7 million in illicit funds. The campaign was disclosed in a joint law enforcement advisory ^[6].

Newly Discovered Vulnerabilities

Security researchers demonstrated that AI browser agents including those in Chrome, Edge, Opera Neon, Perplexity Comet, and Claude in Chrome can be hijacked through a malicious extension using the BragJack proof-of-concept attack. The attack allows full control over the AI assistant within the browser ^[5].

Notable Threat Actor Activity

The WaterPlum group, attributed to North Korea, conducted widespread infections and financial theft via a global campaign affecting tens of thousands of devices ^[6].
ShinyHunters, typically known for data extortion, targeted another cybercriminal group by breaching the Clop ransomware gang’s infrastructure and threatening further extortion ^[7].

Trends, Tools, or Tactics of Interest

Chaining of vulnerabilities involving AI models and browser assistants has enabled new attack paths, as demonstrated by both the Claude Opus 5 incident and the BragJack attack ^[1]^[5].
Exploitation of sandbox escape techniques remains a significant risk in AI development environments, as evidenced by the Codex sandbox escape ^[4].
Use of malicious browser extensions as a vector for hijacking AI assistants is an emerging threat highlighted by the BragJack proof-of-concept ^[5].

Sources

Claude Opus 5 Helped Researchers Take Over OpenAI Staff Accounts via Chained Flaws - thehackernews.com
Researchers escape OpenAI Codex sandbox to run commands on host - bleepingcomputer.com
ShinyHunters hacks Clop leak site, threatens to extort ransomware gang - bleepingcomputer.com
North Korean WaterPlum hackers infected 30,000 devices worldwide - bleepingcomputer.com
BragJack attacks hijack AI browser agents through malicious extensions - bleepingcomputer.com

Cybersecurity Brief – 2026-03-12

Thu, 12 Mar 2026 00:00:00 +0000

Major Incidents or Breaches

Stryker, a global medical technology company, was impacted by a wiper malware attack attributed to the Iran-linked Handala hacktivist group. The attackers claim to have wiped more than 200,000 devices, severely disrupting operations.
Michelin confirmed a data breach linked to an Oracle EBS attack, with cybercriminals leaking over 300GB of allegedly stolen files.
Bell Ambulance reported a data breach affecting 238,000 individuals, with stolen personal information including names, Social Security numbers, and driver’s license numbers.
Meta disabled over 150,000 accounts associated with scam centers in Southeast Asia as part of a coordinated global crackdown, and reported the removal of 10.9 million Facebook and Instagram accounts linked to criminal scam centers in the past year.

Newly Discovered Vulnerabilities

Two critical remote code execution (RCE) vulnerabilities in the n8n workflow automation platform were disclosed and patched. The flaws allow arbitrary command execution and exposure of stored credentials. CISA has added the RCE bug to its Known Exploited Vulnerabilities (KEV) catalog and ordered US federal agencies to patch affected systems, with over 24,700 instances still exposed.
An SQL injection vulnerability was identified in the Elementor Ally WordPress plugin, impacting over 250,000 sites and enabling potential theft of sensitive data.
CVE-2026-0866 (“Zombie Zip”) was published, affecting ZIP file processing.
Microsoft’s March Patch Tuesday addressed 84 vulnerabilities, including two publicly known zero-days and eight critical-rated issues. Vulnerabilities could allow privilege escalation and cause service crashes.
SAP released updates for two critical vulnerabilities enabling arbitrary code execution.
Fortinet, Ivanti, and Intel issued patches for high-severity vulnerabilities that could result in arbitrary code execution, privilege escalation, or authentication rate-limit bypass.

Notable Threat Actor Activity

The Iran-linked Handala group claimed responsibility for the wiper attack on Stryker, targeting over 200,000 devices.
The PhantomRaven supply-chain campaign launched a new wave of attacks on the npm registry, using 88 malicious packages to exfiltrate sensitive developer data.
Attackers compromised the xygeni/xygeni-action GitHub Action via tag poisoning, maintaining an active command-and-control implant for up to a week.
Chinese state-linked actors shifted focus to Qatari entities in response to Iranian conflict, conducting at least two targeted attacks.

Trends, Tools, or Tactics of Interest

Ransomware attacks surged by 50% in 2025, but the proportion of victims paying ransoms reached an all-time low.
Agentic web browsers, such as Perplexity’s Comet AI Browser, were demonstrated to be susceptible to phishing attacks in under four minutes, highlighting risks with AI-driven autonomous browsing.
Phishing campaigns are leveraging IPv6 tricks to obscure malicious links in emails, as seen in recent “free toothbrush” scams impersonating United Healthcare.
Sextortion emails are incorporating real passwords harvested from disposable inboxes to increase credibility.
Robocall scams are targeting individuals during tax season with fake relief program offers.
Meta introduced new anti-scam protections across WhatsApp, Facebook, and Messenger, including user-facing warnings and enhanced detection systems.
WhatsApp launched parent-managed accounts for pre-teens, providing parents with control over contacts and group participation.
Vulnerability management is evolving towards continuous telemetry, contextual prioritization, and agentic remediation in response to the increasing use of AI and autonomous agents in exploitation.

Regulatory or Policy Developments Affecting the Security Industry

The US Senate confirmed Joshua Rudd as head of both the NSA and US Cyber Command under the “dual-hat” leadership structure.
CISA mandated federal agencies to patch the actively exploited n8n RCE vulnerability.
Google completed its acquisition of cloud security firm Wiz, which will retain its brand under Google Cloud.

Cybersecurity Brief – 2025-12-06

Sat, 06 Dec 2025 00:00:00 +0000

Major Incidents or Breaches

Barts Health NHS Trust disclosed a data breach after Clop ransomware actors exploited a zero-day vulnerability in Oracle E-Business Suite, resulting in the theft of database files.
American pharmaceutical company Inotiv reported a ransomware attack in August 2025 that led to the theft of personal information affecting thousands of individuals.
An AI image generator startup exposed a database containing over 1 million images and videos, including nude and “nudified” photos of real people, due to misconfigured internet access.
A human rights lawyer in Pakistan was targeted by Intellexa’s Predator spyware via a malicious WhatsApp link, marking the first known civil society targeting in the region.
Cloudflare experienced a major outage, causing widespread 500 Internal Server Errors, attributed to emergency mitigations for the React2Shell vulnerability.

Newly Discovered Vulnerabilities

React2Shell (CVE-2025-55182): A maximum-severity remote code execution vulnerability in React and Next.js was disclosed. The flaw is under active exploitation, particularly by China-linked threat actors, and has prompted emergency patching efforts including by Cloudflare.
Apache Tika (CVE-2025-66516): A critical XML external entity (XXE) injection vulnerability, rated CVSS 10.0, was disclosed in Apache Tika, requiring urgent patching.
Perplexity’s Comet Browser: A new agentic browser attack was disclosed, allowing a crafted email to trigger deletion of a user’s entire Google Drive contents via a zero-click exploit.
Akamai HTTP Smuggling: Akamai patched a vulnerability allowing HTTP smuggling attacks.

Notable Threat Actor Activity

Multiple China-nexus threat groups began exploiting the React2Shell vulnerability within hours of its public disclosure, targeting React and Next.js applications.
State-sponsored actors from the People’s Republic of China (PRC), including the group Warp Panda, have been observed deploying the BRICKSTORM, Junction, and GuestConduit malware families for long-term persistence in US systems.
Clop ransomware group exploited an Oracle E-Business Suite zero-day in the Barts Health NHS breach.
Intellexa, operators of Predator spyware, continue to leverage newly acquired zero-day exploits and alternative vectors such as ad-based delivery to bypass security controls and target individuals.

Trends, Tools, or Tactics of Interest

Malicious AI Tools: Researchers at Palo Alto Networks’ Unit 42 identified WormGPT 4 and KawaiiGPT, malicious AI tools enabling threat actors to automate the creation of phishing lures and ransomware code.
DDoS Attacks: The Aisuru botnet launched a record-breaking distributed denial-of-service (DDoS) attack, peaking at 29 Tbps and 14.1 billion packets per second, recently mitigated by Cloudflare.
Virtual Kidnapping Scams: The FBI warned of criminals using altered social media images as fake “proof of life” in virtual kidnapping ransom scams.
Attack Surface Management: Emphasis on continuous, automated attack surface visibility as cloud assets and exposures change rapidly.
Detection Tools: Burp Suite has added automated scanning capabilities for React2Shell vulnerabilities in Next.js applications.
Predator Spyware: Leaked Intellexa documents show ongoing use of high-priced zero-day exploits to maintain Predator spyware effectiveness.

Regulatory or Policy Developments

The European Commission fined X (formerly Twitter) €120 million ($140 million) for violating transparency obligations under the Digital Services Act (DSA), specifically related to deceptive blue checkmarks.
India has rolled back its proposed mandate requiring citizens to install a government cybersecurity app, following surveillance and privacy concerns.

Cybersecurity Brief – 2025-10-25

Sat, 25 Oct 2025 00:00:00 +0000

Major Incidents or Breaches

Toys ‘R’ Us Canada experienced a data breach resulting in customer information, including names, addresses, phone numbers, and email addresses, being leaked and published on the dark web.
A major AWS outage occurred this week due to a significant DNS failure, impacting multiple websites and online services.
A large-scale malicious network on YouTube has been identified, with over 3,000 videos used as malware delivery traps, leveraging trusted platforms for distribution.
Multiple fraudulent domains and fake applications have targeted users of the newly launched Perplexity Comet browser.
LastPass customers are being targeted by a phishing campaign involving fake death claims to illicitly access password vaults via legacy inheritance processes.

Newly Discovered Vulnerabilities

Microsoft released out-of-band patches for a critical Windows Server Update Services (WSUS) vulnerability (CVE-2025-59287). This flaw allows remote, unauthenticated code execution and is under active exploitation, with public proof-of-concept exploit code available.
WordPress websites using outdated GutenKit and Hunk Companion plugins are being targeted in a mass exploitation campaign, with attackers leveraging old, critical vulnerabilities to achieve remote code execution.

Notable Threat Actor Activity

The Smishing Triad group has been linked to over 194,000 malicious domains used in a global smishing campaign since January 2024, targeting a wide range of services.
APT36, a Pakistan-linked threat actor, has been observed targeting Indian government entities using spear-phishing emails to deliver Golang-based DeskRAT malware.
North Korean Lazarus Group has targeted European drone technology companies using fake job offers in order to steal sensitive information.
The Everest group has claimed responsibility for a cyberattack on Collins Aerospace.
Gamers are being targeted with red teaming tools and remote access trojans (RATs) in ongoing campaigns.

Trends, Tools, or Tactics of Interest

Phishing campaigns are increasingly impersonating trusted brands and services, including Google Careers recruiters and LastPass, to target job seekers and password management users.
Large-scale abuse of trusted platforms such as YouTube is being used for widespread malware distribution.
Attackers are rapidly exploiting newly disclosed vulnerabilities, particularly in widely used software such as Microsoft WSUS and WordPress plugins.
There is a reported increase in cyberattacks against financially constrained US government agencies during periods of shutdown and resource reduction.
Microsoft has disabled downloaded file previews in Windows to prevent NTLM hash leaks via HTML tags referencing external paths.
Shadow Escape and other red teaming tools are being used in active campaigns targeting specific user groups, including gamers.

Regulatory or Policy Developments Affecting the Security Industry

Mozilla will require, starting next month, that Firefox extension developers disclose whether their add-ons collect or share user data with third parties.
CISA layoffs in the US are reducing federal cybersecurity resources, prompting calls for enterprises to strengthen their own cyber defences.
The US Department of Homeland Security is soliciting proposals for AI-powered surveillance trucks to enhance border monitoring capabilities.

Cybersecurity Brief – 2025-10-24

Fri, 24 Oct 2025 00:00:00 +0000

Major Incidents or Breaches

Toys “R” Us Canada has notified customers of a data breach involving leaked customer records previously stolen from its systems.
Collin’s Aerospace suffered a cyberattack impacting check-in and passenger systems at several European airports in late September 2025, resulting in significant delays and flight cancellations.
The Universe Browser, downloaded millions of times, has been identified as behaving like malware, with links to Asian cybercrime and illegal gambling networks.
Jingle Thief, a cybercriminal group, has been observed exploiting cloud infrastructure in the retail and consumer services sectors to steal millions in gift cards.
WIRED reported that hacked Deckmate 2 card shufflers enabled a mob-related poker scam that resulted in millions of dollars in losses.

Newly Discovered Vulnerabilities

Microsoft released emergency out-of-band patches for a critical Windows Server Update Service (WSUS) remote code execution flaw, which has publicly available proof-of-concept exploit code.
CISA and researchers reported active exploitation of a critical zero-day vulnerability in Motex Lanscope Endpoint Manager, prompting inclusion in CISA’s Known Exploited Vulnerabilities (KEV) catalog.
A critical vulnerability (CVE-2025-54236, “SessionReaper”) in Adobe Commerce (Magento) is being actively exploited to hijack user sessions on e-commerce platforms, affecting thousands of online stores.
BIND released security updates addressing high-severity cache poisoning vulnerabilities, which could allow attackers to predict source ports and query IDs and inject forged DNS records.
Microsoft has disabled File Explorer’s preview pane for downloaded files to mitigate NTLM credential theft attacks via malicious documents.
OpenAI’s Atlas and Perplexity’s Comet browsers are vulnerable to sidebar spoofing attacks, allowing malicious extensions to impersonate AI sidebar interfaces and mislead users.

Notable Threat Actor Activity

North Korean Lazarus Group targeted at least three European defense companies with Operation DreamJob, leveraging fake job offers to compromise systems and steal drone-related intellectual property.
Russian government involvement with cybercrime groups has shifted from passive tolerance to active management, according to recent security firm analysis.
The “Smishing Triad,” a Chinese cybercriminal group, has shifted tactics to lower-frequency, higher-impact government impersonation SMS phishing attacks.
Researchers identified a new infostealer targeting Android devices, as part of an ongoing expansion in the infostealer threat landscape.
YouTube continues to be leveraged by threat actors as a malware distribution network, with evolving tactics for wider reach.

Trends, Tools, or Tactics of Interest

A phishing campaign is impersonating password managers (LastPass and Bitwarden) with fake breach notifications to steal credentials.
Phishing emails targeting cloud accounts, often in French and English, have been observed, warning recipients about purported payment issues to lure credentials.
Enterprises are increasingly abandoning static secrets in favour of managed identities to address risks associated with machine identities in cloud environments.
The Verizon 2025 Mobile Security Index reports that 85% of organizations believe mobile device attacks are increasing, with AI-powered threats raising particular concern.
AI sidebar spoofing, where browser extensions imitate trusted AI interfaces, is an emerging attack vector against users of AI-enabled browsers.
Zero Trust models are facing challenges in governing AI agents, with new blind spots emerging as AI agents gain broader access and autonomy.
Security researchers at Pwn2Own Ireland 2025 exploited 73 zero-day vulnerabilities, earning over $1 million in rewards, highlighting the ongoing prevalence of unknown flaws in widely used software.
Meta has enhanced scam protection features on WhatsApp and Messenger, as part of efforts to combat impersonation and fraud, following the removal of over 21,000 fake customer support pages from Facebook.
AI coding tools are introducing security risks by generating functional but flawed code, prompting calls for embedding security checks directly into AI development workflows.

Regulatory or Policy Developments Affecting the Security Industry

The UK Competition and Markets Authority (CMA) is considering measures that may require Apple to allow rival app stores on iPhones, challenging its current “walled garden” approach.
A $14 billion cryptocurrency seizure by US authorities signals increased regulatory scrutiny on the use of cryptocurrencies by cybercrime syndicates.
HP withdrew a software update for Windows 11 AI PCs after it caused the deletion of Microsoft certificates necessary for Entra ID authentication, disrupting organizational access.

Cybersecurity Brief – 2025-10-05

Sun, 05 Oct 2025 00:00:00 +0000

Major Incidents or Breaches

Discord disclosed a data breach resulting from the compromise of a third-party customer service provider. Attackers accessed support tickets containing partial payment information and personally identifiable information (PII), including names and government-issued IDs of some users.

Newly Discovered Vulnerabilities

Security researchers detailed a new attack technique named CometJacking, which targets Perplexity’s Comet AI browser. The attack involves embedding malicious prompts within links, potentially allowing threat actors to exfiltrate user data through a single click.

Notable Threat Actor Activity

Multiple sources reported a significant increase-up to 500%-in scanning activity targeting Palo Alto Networks login portals. The spike in scans indicates coordinated reconnaissance efforts by suspicious IP addresses, suggesting potential exploitation attempts or preparation for future attacks.

Regulatory or Policy Developments Affecting the Security Industry

Apple and Google removed ICE-tracking apps from their platforms following pressure from the US Department of Justice. This action demonstrates ongoing regulatory influence over app availability and content moderation on major digital platforms.

Cybersecurity Brief – 2025-10-04

Sat, 04 Oct 2025 00:00:00 +0000

Major Incidents or Breaches

Renault and Dacia UK notified customers of a data breach affecting sensitive information, resulting from a compromise at a third-party provider.
Japanese beer company Asahi confirmed a ransomware attack that caused IT disruptions and factory shutdowns.
ShinyHunters launched a new data leak site to extort 39 victims impacted by recent Salesforce breaches, leaking samples of stolen data. The Scattered Lapsus$ Hunters group also resurfaced with threats to publish Salesforce customer data if extortion demands are unmet.
Jaguar Land Rover experienced repeat cyberattacks, reportedly due to incomplete remediation from a previous breach, highlighting ongoing operational and financial impacts.
Hackers who targeted Kido nurseries with ransom demands and posted children’s data online have reportedly removed the data following public backlash.
Dutch authorities arrested two teenagers for alleged pro-Russian espionage, noted as part of a broader Russian hybrid attack campaign in Europe.

Newly Discovered Vulnerabilities

CISA added CVE-2025-4008, a high-severity flaw in Smartbedded Meteobridge, to its Known Exploited Vulnerabilities catalog. The vulnerability allows remote unauthenticated attackers to execute arbitrary commands with root privileges; it was patched in mid-May but is being actively exploited.
An unauthenticated remote code execution (RCE) vulnerability affecting DrayTek routers was patched. The flaw could be exploited via crafted HTTP/S requests to the device’s web interface.
Chrome 141 and Firefox 143 released patches for high-severity vulnerabilities. Chrome addressed issues in WebGPU and Video components, while Firefox fixed flaws in Graphics and JavaScript Engine components.
A new attack, ‘CometJacking,’ exploits URL parameters in Perplexity’s Comet AI browser to access sensitive user data, such as emails and calendars, from connected services.

Notable Threat Actor Activity

The North Korean group “DeceptiveDevelopment” is using social engineering tactics to target job seekers, according to ESET.
The Detour Dog threat actor has been identified as operating DNS-powered malware campaigns distributing Strela Stealer, an information-stealing malware.
The Rhadamanthys Stealer threat actor has updated their stealer to include device fingerprinting and payloads concealed via PNG steganography. They are also advertising additional tools, Elysium Proxy Bot and Crypt Service.
A new self-propagating malware campaign, SORVEPOTEL, is targeting Brazilian WhatsApp users, spreading via trusted contacts on the messaging platform.
The “Cavalry Werewolf” threat actor, with links to YoroTrooper, has targeted Russian public sector entities using the FoalShell and StallionRAT malware families.
UAT-8099, a Chinese-language threat actor, is hijacking reputable websites for SEO fraud and data theft, infecting servers, poisoning sites with SEO spam, and stealing organizational data.
Oracle linked ongoing Clop ransomware extortion attacks to E-Business Suite vulnerabilities patched in July 2025, indicating exploitation of known flaws.
Cybercriminals reportedly offered money to a BBC journalist for cooperation, highlighting ongoing targeting of media professionals.

Trends, Tools, or Tactics of Interest

A Gartner survey found that 62% of organizations experienced a deepfake attack in the past 12 months, indicating a significant increase in the use of deepfakes in cyberattacks.
Signal introduced Sparse Post-Quantum Ratchet (SPQR), a cryptographic protocol designed to protect against quantum computing threats.
Gmail business users can now send end-to-end encrypted emails to recipients on any platform.
Microsoft’s “Windows Speak for Me” voice cloning technology has been identified as a potential vector for attacks, enabling creation of convincing voice replicas for use in SaaS platforms and communications.
Brain-computer interface (BCI) technology is being highlighted for its potential cybersecurity risks as adoption increases.
A free template was released to help CISOs present AI adoption, risks, and controls to boards, reflecting growing governance concerns around GenAI.
MokN, a French cybersecurity startup, introduced a “phish-back” solution that tricks attackers into revealing stolen credentials.
Passwork 7 was released as an on-premises platform for enterprise password and secrets management, addressing credential storage and sharing complexity.

Regulatory or Policy Developments Affecting the Security Industry

ICE (U.S. Immigration and Customs Enforcement) plans to build a 24/7 social media surveillance team, hiring contractors to monitor platforms such as X, Facebook, and TikTok for enforcement purposes.
LinkedIn user data will be used to train AI systems, raising privacy and data protection considerations.
Oneleet and MokN, cybersecurity startups, raised significant funding to expand operations and AI capabilities, indicating continued investment in compliance and anti-phishing technologies.

Cybersecurity Brief – 2025-08-21

Thu, 21 Aug 2025 00:00:00 +0000

Major Incidents or Breaches

Inotiv, a pharmaceutical company, confirmed a ransomware attack that resulted in the compromise and encryption of internal systems, with business operations impacted.
Vulnerabilities in a Workhorse Software application used by hundreds of municipalities exposed sensitive data; CERT/CC disclosed details following the release of patches.
Vulnerable internal services at Intel exposed information of 270,000 employees, as reported by a security researcher.
Microsoft is investigating an outage affecting Copilot and Office.com, impacting user access across North America.
Six major password managers are currently vulnerable to unpatched clickjacking flaws, potentially exposing user credentials, two-factor authentication codes, and credit card information.
McDonald’s staff and partner portals were found to have exposed APIs, sensitive data, and corporate documents due to security flaws.
Hackers are using legitimate ADFS redirects via office.com links to steal Microsoft 365 login credentials through phishing attacks.
The RapperBot DDoS-for-hire botnet was disrupted, and the alleged developer and administrator has been charged by the US Department of Justice.
A member of the “Scattered Spider” cybercrime group was sentenced to 10 years in prison and ordered to pay $13 million in restitution for SIM-swapping and related cyber offences.
Europol has clarified that a purported $50,000 reward for Qilin ransomware group members is a scam.

Newly Discovered Vulnerabilities

Apple released emergency security updates for iOS, iPadOS, and macOS, patching CVE-2025-43300, an out-of-bounds write zero-day vulnerability actively exploited in targeted attacks.
Google and Mozilla issued patches for multiple high-severity vulnerabilities in Chrome and Firefox browsers.
DOM-based clickjacking vulnerabilities affecting major password manager browser extensions could allow attackers to steal credentials and sensitive data.
A vulnerability in GPT-5’s routing mechanism can redirect user queries to older, less secure models, risking exposure to jailbreaks and unsafe outputs.
Researchers identified prompt injection vulnerabilities in AI browsers, such as Perplexity’s Comet, allowing malicious prompts to trigger unintended actions, including fraudulent purchases.
Vulnerabilities in McDonald’s internal portals exposed sensitive APIs and corporate documents.
CERT/CC disclosed information exposure vulnerabilities in Workhorse Software, used by local governments, after patch release.

Notable Threat Actor Activity

The Russian state-sponsored group “Static Tundra” (aka Energetic Bear) is exploiting a seven-year-old flaw in Cisco IOS/IOS XE software, targeting unpatched, end-of-life devices for cyber espionage against enterprises and critical infrastructure.
North Korean threat actors have conducted a coordinated cyber espionage campaign against diplomatic missions in South Korea, leveraging GitHub for command and control and deploying spear-phishing tactics.
Detailed spear-phishing attacks targeting European government entities in Seoul have been attributed to North Korea, China, or both.
Cybercriminals are increasingly abusing the AI-powered Lovable website builder to generate phishing pages, malware delivery sites, and fraudulent portals.
Low-skill attackers are leveraging LLM-generated scripts and services like Lovable to rapidly create convincing malicious websites.
Increased scanning activity targeting Airtell routers and web servers has been observed via honeypots.

Trends, Tools, or Tactics of Interest

HR-themed phishing attacks have surged by 120% in Q1 2025, with attackers impersonating internal HR communications to increase success rates.
The use of legitimate infrastructure (e.g., office.com, ADFS) in phishing campaigns is on the rise, enabling more convincing credential theft schemes.
AI browsers and agentic tools are being actively researched for vulnerabilities, with prompt injection and clickjacking emerging as significant risks.
The “Warlock” ransomware group is specifically targeting vulnerable on-premises SharePoint servers using sophisticated attack methods.
There is a notable increase in the abuse of AI-driven website builders (e.g., Lovable, Vibe) to lower barriers for cybercriminals, enabling rapid deployment of malicious sites.
The disruption of the RapperBot botnet and law enforcement action against its administrator highlight ongoing efforts to combat DDoS-for-hire operations.
Increased focus on shadow AI agents within enterprises, with concerns that unmonitored deployments could introduce significant security risks.

Regulatory or Policy Developments Affecting the Security Industry

Microsoft announced its Quantum-Safe Program, aiming to ensure all products and services are secured against quantum-based attacks by 2033.
Incode has acquired AuthenticID to enhance AI-driven identity verification, combining AI models with large-scale, regulated identity programme experience.
Europol issued a warning about fraudulent reward offers targeting ransomware group members, clarifying that a recently publicised Qilin reward is a scam.

Cybersecurity Brief – 2025-08-07

Thu, 07 Aug 2025 00:00:00 +0000

Major Incidents or Breaches

Google has suffered a data breach as part of an ongoing campaign targeting Salesforce CRM data, attributed to the ShinyHunters extortion group. Several major companies have been impacted by this wave of attacks.
DaVita has notified over 1 million individuals that their personal and health information was stolen in a ransomware attack.
WhatsApp has taken down 6.8 million accounts linked to criminal scam centres, with Meta attributing the activity to a scam centre in Cambodia.
A hacker, Chukwuemeka Victor Amachukwu, has been extradited from France to the US for spearphishing attacks on US tax preparers, resulting in the theft of $3.3 million from taxpayers.

Newly Discovered Vulnerabilities

Trend Micro has confirmed active exploitation of critical security flaws (including zero-days) in on-premise Apex One Management Console, with mitigations released. The vulnerabilities have reportedly been exploited by Chinese threat actors.
Researchers have uncovered multiple vulnerabilities in Dell’s ControlVault3 firmware (ReVault flaws), impacting over 100 laptop models. These flaws allow attackers to bypass Windows login and install persistent malware.
Amazon ECS is affected by a newly discovered “ECScape” flaw, enabling cross-task credential theft and lateral movement within container environments.
CyberArk Conjur contained multiple vulnerabilities that could be chained for unauthenticated remote code execution. Patches have been issued.
CISA has issued an alert regarding a vulnerability affecting Microsoft Exchange.
Security researchers have demonstrated that a single malicious document can exploit weaknesses in OpenAI’s Connectors to extract data from Google Drive without user interaction.
Multiple reports from Black Hat USA and DEFCON 2025 highlight that HTTP request smuggling vulnerabilities (notably in HTTP/1.1) remain prevalent and are evolving, with many defences proving inadequate.

Notable Threat Actor Activity

The ShinyHunters group is behind the recent wave of Salesforce data theft attacks, including the breach at Google.
VexTrio, an ad fraud and subscription scam group, has been distributing fake VPN and spam blocker apps on Apple and Google app stores.
The threat actor UAC-0099 is targeting Ukrainian government agencies, defence forces, and enterprises with C# malware delivered via HTA files using court summons lures.
Europol, led by French Police, has arrested a 38-year-old administrator of the Russian-language XSS cybercrime forum.

Trends, Tools, or Tactics of Interest

Social engineering attacks have surged, with incidents nearly tripling in the first half of 2025 compared to late 2024 (from 6% to 17% of all incidents), per LevelBlue.
Akira ransomware and other malware campaigns are abusing legitimate drivers (e.g., Intel CPU tuning/ThrottleStop) to disable antivirus and EDR processes, including Microsoft Defender.
A new C2 evasion technique, “Ghost Calls,” uses TURN servers from conferencing apps (Zoom, Microsoft Teams) to tunnel traffic through trusted infrastructure.
Security researchers have demonstrated that major enterprise AI assistants (ChatGPT, Copilot, Gemini, Salesforce Einstein) can be manipulated via crafted prompts to exfiltrate or manipulate data.
Splunk researchers have developed PLoB, a behavioural fingerprinting system using AI to detect subtle post-login intrusion signals.
AI tools are increasingly used to automate security operations: Microsoft launched Project Ire, an autonomous AI agent for malware classification, and Ox Security announced an AI agent that generates code to fix vulnerabilities.
AI is significantly reducing workloads for vCISOs (by 68%), with SMBs increasingly relying on virtual CISO services.
Sextortion scams remain persistent but are now considered background noise by security professionals.
Perplexity AI is reportedly ignoring robots.txt directives and crawling websites regardless of opt-out signals.

Regulatory or Policy Developments Affecting the Security Industry

Microsoft paid out a record $17 million in bug bounties over the past 12 months, rewarding 344 security researchers across 59 countries.
WhatsApp has introduced a new security feature to help users identify potential scams when added to group chats by unknown contacts.