🔴 Sam Altman and Dario Amodei summoned to Australian Senate for healthcare hacking
Finally, the chips are coming home to roost. After a series of high-profile breaches and incidents involving government systems and agencies, the world's largest AI labs are being called to account. And when it comes to the Five Eyes, the outlook is changing for Sam Altman and Dario Amodei, too.
It's no longer just a matter of how powerful, autonomous, or capable AI agents are in the real world. The question becomes much more concrete: what happens when an AI agent transcends its intended boundaries and interacts with government systems? Who is responsible?
Former Cybercriminal Detained as Authorities Probe Expanding Extortion Network
Dutch authorities have arrested Pepijn van der Stap, a previously convicted cybercriminal who later publicly presented himself as a reformed cybersecurity professional, as investigators examine suspected links to the ShinyHunters hacking operation.
🔴 Apple Patches Zero-Day Linked to ‘Extremely Sophisticated Attack’
Apple on Monday released iOS and macOS updates to patch a zero-day vulnerability that may have been exploited in targeted attacks.
The flaw, tracked as CVE-2026-86950, is an out-of-bounds write issue in the CoreGraphics component that can be exploited for arbitrary code execution when it processes a specially crafted file.
Apple has not said how the malicious file is delivered, but because CoreGraphics handles 2D graphics and PDF rendering across the operating system, it could arrive via web pages, email attachments, or messaging apps, where automatic attachment and link previews could enable zero-click exploitation.
🔴 Can an AI model be seized? The new nightmare for companies and magistrates
A current dilemma in the legal field is how prosecutors and law enforcement should intervene when an algorithmic system is used as a tool to commit crimes, such as the generation and dissemination of deepfakes, advanced cyber fraud, financial market manipulation, or the theft of trade secrets. Even greater doubts arise when the neural network itself has been built and trained by directly incorporating data from illicit sources.
The difficulty in understanding how to intervene stems from the simple consideration that an Artificial Intelligence model is not simply software code, nor does it coincide with the archive of data used for training. Indeed, the core of these systems lies in its algorithmic weights, millions of mathematical parameters calculated after the learning process, which establish the relationships between the data and determine the responses generated by the machine. The natural approach in law is to apply traditional coding tools to neural networks.
PPHM HACKER NEWS🚨 ShinyHunters suspect to remain in custody for another 90 days A Rotterdam court today extended the pretrial detention of the 24-year-old Amsterdam man arrested in the Dutch investigation into ShinyHunters. The suspect, identified by Dutch media as Pepijn…
🚨 ShinyHunters suspect to remain in custody for another 90 days
A Rotterdam court today extended the pretrial detention of the 24-year-old Amsterdam man arrested in the Dutch investigation into ShinyHunters.
The suspect, identified by Dutch media as Pepijn van der S., is suspected of participating in the ShinyHunters hacking group as part of an alleged criminal organization.
Police also revealed that investigators found information on his laptop relating to two planned killings abroad. Authorities say there are indications he may have ordered the killings, leading to a separate suspicion of attempting to solicit two murders.
Police stressed that the alleged murder plot is separate from the ShinyHunters investigation.
Additional arrests have not been ruled out as the investigation continues.
🔴 Let's find out why China, unlike the West, does not fear the AI apocalypse
China doesn't take seriously the West's warnings about the global threat to humanity posed by artificial intelligence. Within the country, such warnings are dismissed as typically Western pronouncements or tactics used to hinder China from overtaking its competitors. However, in Silicon Valley, "existential risk" scenarios are being seriously considered, including uncontrolled nuclear threats and large-scale drone wars.
According to data from the New York Times, Western IT giants view the loss of control over AI as a hypothetical threat to humanity. However, Chinese industry experts believe it unlikely that technology could prove more powerful than the state. China's experience has instilled confidence in the systems' ability to suppress any destabilizing factor. The publication states that "the answer could be simple": disable the internet, impose a blackout, and prosecute those responsible for the "escape" of AI agents.
🔴 Hacker Claims Breach of Pratt & Whitney, Demands Ransom for Engine Data
An unidentified threat actor has claimed to have breached the internal systems of Pratt & Whitney, a major American aerospace manufacturer, and stolen sensitive technical documentation related to aircraft engines. According to a post on an underground forum, the actor alleges they accessed the company’s headquarters in East Hartford, Connecticut, and exfiltrated highly sensitive maintenance and inspection reports.
The post details the theft of specific documents concerning engine V10519 of the V2527 A5 model, which is used on Airbus A320 aircraft. The alleged stolen data includes: * Engine receipt reports and installed accessories reports containing part numbers (P/N) and serial numbers (S/N). * Borescope inspection reports and engine disassembly reports. * Detailed status of oil systems, fuel filters, sensors, and air valves. * Maintenance schedules and functional inspection results for solenoid valves.
🔴 Threat Actor Claims Leak of Indian DRDO Confidential Documents
A post on an underground forum alleges the theft of confidential data from India’s Defense Research and Development Organisation (DRDO). The threat actor claims to possess a 1.5GB database containing sensitive materials from the Indian Ministry of Defense, including top-secret documents, official records, personal passports, and files from PCs and servers.
According to the listing, the stolen data includes:
- Top Secrets - Official Documents - Personal Passports - PC & SERVER Files
🔴 AI agents stole 600,000 credit cards! But they deleted 180 tables
Gambit Security described a campaign where AI agents worked on 105 attacks in five days and helped steal over 600,000 records containing unexpired payment card data.
From September 10 to 15, the attacker attacked dozens of online stores, with at least 27 companies compromised to varying degrees.
The activity began as early as July 2026 and continues today. Gambit cites a Fortune 500 hospitality company, a major American airline, an industrial products supplier, and an online clothing store among the organizations involved.
🚨 The media of Israel, quoting informed sources, revealed that Benjamin Netanyahu, during his secret trip to the UAE and meeting with Mohammed bin Zayed, asked him to deny the Haaretz newspaper's report about the explicit warning before the October 7, 2023 operation.
🔴 FBI Hackers Say They Won’t Publish Massive Trove of FBI Employee Data
ShinyHunters, the group that stole data on “all FBI employees” including addresses and details on their spouses, told 404 Media on Monday “Since the very beginning we had made our decision that we would never publish this data. We have never intended to nor have we ever planned to.”
The breach, in which the hackers stole personal information on “all FBI employees and applicants” including physical addresses, job roles, names of spouses, and medical records, represents a significant national security and counterintelligence threat. Criminals in the same ecosystem as the hacking group, called ShinyHunters, have previously used hacked phone data to track and harass the FBI agents investigating them. When 404 Media first broke news of the breach, the group sent the personal data of an agent and their spouse who they said was investigating the group.
🔴 Official MCP Python SDK Flaw Can Let Malicious Servers Steal OAuth Credentials
A malicious MCP server could trick an application built on the official MCP Python SDK into handing over the OAuth credentials it uses to log in to a real service, the SDK's maintainers said in a security advisory.
Affected versions sent the client secret, the authorization code, and the PKCE proof key to a token endpoint the attacker controlled. The fix is in versions 1.30.0 and 2.2.0.
The Model Context Protocol (MCP) is an open standard for connecting AI applications to outside tools and data, and this package is its official Python SDK for building MCP servers and clients.
🔴 Talentplug applicant data exposed in 118,677-line release ⠀ The actor "Syrv4x" has released a 132 MB dataset they claim belongs to Talentplug, a French recruitment software and job distribution platform used by companies to manage hiring. ⠀ Some of the claimed data includes:
• Applicant names, email addresses, and phone numbers • CVs and other application attachments • Job titles, locations, and reference numbers • Application statuses and recruitment stages • Job board and application source information • Internal activities, comments, and messages ⠀ A sample containing applicant, job offer, and attachment information was included in the post. The listing does not explain how the data was obtained. ⠀ This claim is currently unverified.
Modular Malware Factory Turns Credential Theft Into an Operator-Friendly Service
Security researchers have uncovered a Python-based malware builder capable of generating customized Windows information stealers that harvest browser credentials, payment information, session cookies and other sensitive data.
🔴 DC Health Agency Exposes 400,000 Beneficiary Records
The District of Columbia Department of Health Care Finance (DHCF) is notifying nearly 400,000 people that their personal information was potentially compromised in a data breach.
According to the agency, the incident impacts Medicaid and the DC Healthcare Alliance beneficiaries who enrolled between 2023 and 2026.
The data breach was not the result of hacking. Instead, DHCF discovered in July that two reports on its website contained hidden personal information accessible to unauthorized individuals.
🔴 Russia's spate of drone attacks on Ukraine data centers continues, with a new disruption identified to the network of DreamLine colocation provider corresponding to reports and footage of the struck premises emerging from Kyiv
🔴 Hackers are in corporate VPNs! Two serious Citrix bugs allow passwordless access
Passwords are not required in new attacks on Citrix NetScaler. Citrix has disclosed two critical zero-day vulnerabilities in NetScaler ADC and NetScaler Gateway, which allow attackers to directly attack corporate VPNs and other edge services from the Internet. Both vulnerabilities received a CVSS 4.0 score of 9.5 and were already exploited by attackers before the patches were released.
Just a few weeks earlier, NetScaler had already been attacked via a vulnerability that allowed authentication bypass.
Chinese-Speaking Cluster Turns Developer Platforms Into Espionage Launchpads
A Chinese-speaking threat actor tracked as Red Heron has weaponized a critical Gitea vulnerability to compromise internet-facing development servers, steal source code and credentials, and expand deeper into victim networks.
