databreaches.net - Chatox and Brosix are communications platforms that advertise for personal use and team use. They are owned by Stefan Chekanov.
The only statement Chatox makes about its data security is “Chatox employs encryption across all communications, making it an extremely secure communication and collaboration platform.”
Brosix Enterprise advertises its security:
Brosix provides you with an efficient and secure communication environment, and Text Chat is a central element of this. With this feature you can instantly send, and receive, text messages to your network contacts. Better yet, all messages sent with Brosix are fully encrypted using end-to-end encryption technology, guaranteeing that your communication remains secure.
Brosix uses AES (Advanced Encryption Standard, used by US government) with 256 bit keys. Which means the encryption can’t be broken in a reasonable time.
All communication channels are direct, peer-to-peer, between the users and are not routed through Brosix servers. In some cases, if user firewalls do not allow direct connection, data is routed through Brosix servers. In these rare cases, the channels through the servers are built in a way that Brosix cannot decrypt and see the user data that flows.
So why did a researcher find a lot sensitive chats in plain text with individuals’ first and last names, username, password, IP address, chat message, and attached files — all unencrypted?
What to Know
A researcher contacted DataBreaches after finding an unsecured backup with 155.3 GB of unique compressed files.
There was a total of 980,972 entries in the users’ tables, with entries going back to 2006.
The researcher first logged the backup as exposed in late April. From the logs, the researcher stated that the files in question were exposed from at least May 11th 2024 – July 4th 2025 . Because logging only began in late April, the server could have been exposed before then.
The top email domains for each of the two platforms are listed below:
Brosix Enterprise Database Chatox Database
14826 gmail.com
5472 yahoo.com
2086 hotmail.com
1805 mail.ru
1111 allstate.com
679 rankinteractive.com
633 yandex.ru
582 issta.co.il
376 outlook.com
353 gp-servicedirect.com 63291 mail.ru
48075 gmail.com
20099 yandex.ru
13789 yahoo.com
7868 hotmail.com
6734 bk.ru
4541 allstate.com
3316 rambler.ru
3297 inbox.ru
3204 list.ru
san.com - Data stolen by a ransomware gang has exposed highly sensitive information from a Louisiana sheriff’s office, including the names, telephone numbers and Social Security numbers of confidential informants in criminal investigations. Straight Arrow News obtained a copy of the data from DDoSecrets, a non-profit that archives hacked and leaked documents in the public interest.
Medusa, a suspected Russian cybercrime group, said on its Dark Web blog in April 2024 that it had pilfered more than 90 gigabytes of data from the East Baton Rouge Sheriff’s Office.
The sheriff’s office initially claimed the intrusion had been quickly detected and stopped, allowing the hackers to obtain only a limited amount of data, such as “screenshots of file folders and still images from video files, WBRZ-TV reported.
65,000 files
A sample of the stolen files shared at the time by Medusa included payroll information, showing that the breach was more substantial than first claimed by the sheriff’s office. Medusa threatened to release all of the data, which contains over 65,000 files, unless the sheriff’s office paid $300,000. There’s no indication the ransom was ever paid.
The East Baton Rouge Sheriff’s Office did not respond to a request for comment from SAN.
SAN’s analysis of the full data cache provides an insight into just how damaging the breach was. Given the sensitivity of the data, DDoSecrets is only sharing it with approved journalists, researchers and defense attorneys practicing in Baton Rouge.
The data covers both the banal day-to-day operations of a law enforcement agency and the potentially life-and-death details of drug cases and other criminal investigations.
“The East Baton Rouge Sheriff’s Office data is an extraordinary example of the inner workings of a police department, down to Internal Affairs investigations and details about the use of confidential informants,” DDoSecrets co-founder Emma Best told SAN. “While the police are obviously of public interest and deserve no privacy, their targets and victims do. With that in mind, we’re refraining from republishing the full data to the public while encouraging journalists and civil rights advocates to engage with it.”
Best said the data cache was posted by Medusa to the messaging app Telegram, but that their channels were repeatedly shut down. The contents of the breach have not been extensively reported on until now.
Law enforcement entities are common targets for ransomware gangs. In 2021, the Metropolitan Police Department in Washington, D.C., was hacked by a Russian-speaking ransomware group known as Babuk, resulting in the leak of 250 gigabytes of data after the department refused to pay a ransom. The data also included sensitive information on informants and police officers.
Confidential informants
Contracts signed by 34 confidential informants in 2023 are among the exposed data from Louisiana.
A document titled “CI Information” lists the names, dates of birth and Social Security numbers of 200 confidential informants involved in narcotics investigations. Names of deputies overseeing informants and case numbers are included, as well as whether the informants are still active. Deactivation dates, indicating when an informant’s work ended, range from 2020 to 2023.
A folder titled “C.I. G.P.S. routes” contains numerous images of maps detailing the movements of informants across Baton Rouge.
Seized devices
A document last edited in August 2023 lists devices seized by the sheriff’s office, primarily mobile phones. The document notes whether a warrant had been requested or obtained, as well as additional steps that may have been needed to access a device’s contents.
Several phones were turned over to the FBI, the data indicates. Some files mention that cellphone hacking tools were needed to pull data from the devices. Files refer to both Cellebrite, an Israeli company that produces tools for extracting data from mobile devices, and GrayKey, a mobile forensics tool developed by the US-based company Grayshift that similarly unlocks and extracts data from phones.
The data also shows that the Drug Enforcement Agency sought access to historical location data and other information from a target’s cell phone.
Cell phone surveillance
Pen trap and trace search warrants — court orders that allow law enforcement to collect cell phone metadata such as numbers dialed — were issued to cellular service providers T-Mobile, AT&T and Verizon.
Many of the warrants mention the use of a “cell site simulator,” also known as an IMSI catcher, to reveal a suspect’s whereabouts. Cell site simulators, commonly referred to as Stingrays, are devices that mimic cell phone towers and can be used to pinpoint the location of specific phones.
Sock puppet accounts
A presentation about online investigations advises officers to create “sock puppet accounts,” a term used to describe a false online identity created to conceal an individual’s real one.
For instance, deputies were told to use a free VPN browser add-on for Google Chrome to hide their IP addresses. The website thisxdoesnotexist.com is also listed as a resource for deputies to create AI-generated images of everything from fake people to resumes.
Hidden cameras and drones
A folder titled “Tech” includes brochures listing an array of surveillance technology, such as GPS trackers and hidden cameras that can be placed inside items such as clothing, vape pens and Newport menthol cigarette packs.
A list of hidden cameras contains IP addresses, login credentials for remote access and identifying information for both the devices and SIM cards used.
One list shows 19 drones operated by the sheriff’s office, the majority of which are made by the Chinese manufacturer DJI. The drones are used by several divisions of the sheriff’s office, including SWAT and narcotics, for suspect apprehension and search and rescue missions.
A PowerPoint presentation in the data cache shows the default password used to access the internal system for logging drone usage. A folder titled “Operation Photos & Videos” shows both surveillance of criminal suspects as well as overhead images of sheriff’s deputies at a shooting range.
Internal affairs
Internal affairs data, including complaints made against the sheriff’s office, accuse deputies of racial profiling, unwarranted searches and excessive force.
Incidents range from a deputy being reprimanded for letting his 10- and 12-year-old children drive his patrol vehicle to another being arrested for battery and suspended for 30 days after being involved in a “road rage-type” episode.
Polygraph results
Other files detail the results of polygraph tests given to both deputies and suspects.
One file graphically details an alleged sexual assault and concludes that the person being tested had been deceitful. A deputy was also accused of being deceitful after being asked whether he’d referred to homosexuals as “disgusting” when discussing a fellow deputy believed to be gay.
Bern, 29.07.2025 — The Office of the Attorney General of Switzerland (OAG) has been conducting criminal proceedings since 2022 in the matter of a large-scale phishing series. Fake e-banking login pages had been used to defraud numerous Swiss bank customers, resulting in losses of around CHF 2.4 million. In this context, the OAG took over about thirty cases from the cantons. The investigations conducted by the OAG and fedpol led to the identification and location of the developer and distributor of phishing kit in the UK. The case was taken over by the British authorities, who were already conducting similar proceedings against the individual involved. He was sentenced by a court in the UK on 23 July 2025 to seven years imprisonment. This success demonstrates the importance of international cooperation in the fight against cybercrime.
In July 2022, the Office of the Attorney General of Switzerland (OAG) initiated criminal proceedings against persons unknown on suspicion of computer fraud (Art. 147 para. 1 in conjunction with para. 2 Swiss Criminal Code (SCC)) in connection with an extensive phishing series. Prior to this, several cantonal public prosecutor's offices had already initiated proceedings in around 30 cases in connection with the same matter, which the OAG subsequently took over and joined in its proceedings. In August 2023, following the identification of the developer and distributor of the phishing kit, criminal proceedings were extended to this person.
Real-time phishing on a grand scale
Between May 2022 and September 2022, unknown perpetrators created and used several fake login websites (phishing pages) for various Swiss banks, using what is known as a phishing kit. Bank customers who used Google Search to access their account ended up on the phishing pages posted as adverts and fell victim to the scam when they attempted to log into their supposed e-banking accounts. As a result, their e-banking access data were intercepted unbeknown to them, enabling the perpetrators to use the stolen access data to log into the victim's e-banking accounts and enable the two-factor authentication. The victims still believed that they were on the bank's real website and authenticated the login by entering the authentication code they received by text message on the phishing page. As a result, the perpetrators gained access to their authentication codes. This enabled them to successfully log into the victims' e-banking accounts and register an additional device with the bank to confirm two-factor authentication. The perpetrators were then able to log into the victims’ e-banking accounts without any further action by the victims and initiate payments without their knowledge or consent. The damage caused to the injured parties in the Swiss criminal proceedings amounts to CHF 2.4 million.
Successful cooperation with the UK, Europol and Eurojust
The intensive investigations conducted by the OAG and fedpol resulted in the identification and localisation of a British national who had developed and distributed the phishing kit. The OAG and fedpol's subsequent close cooperation with Europol, Eurojust and UK law enforcement authorities led to the arrest and prosecution in the UK of the developer and seller of the phishing kit. As the UK authorities were already conducting similar proceedings against this person, they took over the Swiss proceedings at the OAG’s request, continuing them in the UK. The OAG subsequently discontinued its criminal proceedings. On 23 July 2025, the perpetrator was sentenced in the UK to seven years imprisonment for his offences (press release from the Crown Prosecution Service). This success demonstrates the importance and effectiveness of international cooperation in tackling the fight against the ever-increasing cybercrime.
thedfirreport.com - Bumblebee malware has been an initial access tool used by threat actors since late 2021. In 2023 the malware was first reported as using SEO poisoning as a delivery mechanism. Recently in May of 2025 Cyjax reported on a campaign using this method again, impersonating various IT tools. We observed a similar campaign in July in which a download of an IT management tool ended with Akira ransomware.
In July 2025, we observed a threat actor compromise an organization through this SEO poisoning campaign. A user searching for “ManageEngine OpManager” was directed to a malicious website, which delivered a trojanized software installer. This action led to the deployment of the Bumblebee malware, granting the threat actor initial access to the environment. The intrusion quickly escalated from a single infected host to a full-scale network compromise.
Following initial access, the threat actor moved laterally to a domain controller, dumped credentials, installed persistent remote access tools, and exfiltrated data using an SFTP client. The intrusion culminated in the deployment of Akira ransomware across the root domain. The threat actor returned two days later to repeat the process, encrypting systems within a child domain and causing significant operational disruption across the enterprise.
This campaign affected multiple organizations during July as we received confirmation of a similar intrusion responded to by the Swisscom B2B CSIRT in which a malicious IT tool dropped Bumblebee and also ended with Akira ransomware deployment.
The Wiz Research team has discovered a chain of critical vulnerabilities in NVIDIA's Triton Inference Server, a popular open-source platform for running AI models at scale. When chained together, these flaws can potentially allow a remote, unauthenticated attacker to gain complete control of the server, achieving remote code execution (RCE).
This attack path originates in the server's Python backend and starts with a minor information leak that cleverly escalates into a full system compromise. This poses a critical risk to organizations using Triton for AI/ML, as a successful attack could lead to the theft of valuable AI models, exposure of sensitive data, manipulating the AI model's responses and a foothold for attackers to move deeper into a network.
Wiz Research responsibly disclosed these findings to NVIDIA, and a patch has been released. We would like to thank the NVIDIA security team for their excellent collaboration and swift response. NVIDIA has assigned the following identifiers to this vulnerability chain: CVE-2025-23319, CVE-2025-23320, and CVE-2025-23334. We strongly recommend all Triton Inference Server users update to the latest version. This post provides a high-level overview of these new vulnerabilities and their potential impact.
The enclosed work is the latest in a series of NVIDIA vulnerabilities we’ve disclosed, including two container escapes: CVE-2025-23266 and CVE 2024-0132.
Mitigations
Update Immediately: The primary mitigation is to upgrade both the NVIDIA Triton Inference Server and the Python backend to version 25.07 as advised in the NVIDIA security bulletin.
Wiz customers can use the following to detect vulnerable instances in their cloud environment:
Wiz customers can use the Vulnerability Findings page to find all instances of these vulnerabilities in their environment, or filter results to instances related to critical issues. Alternatively, you can use the Security Graph to identify publicly exposed vulnerable VMs/serverless or containers.
Wiz Advanced customers can filter on findings identified or validated by the Dynamic Scanner, and Wiz Sensor customers can filter on runtime validated findings.
Wiz Code customers can filter on findings with one-click remediation to generate a pull request with fixes for vulnerable instances detected in code repositories.
therecord.media - Multiple cybersecurity incident response firms are warning about the possibility that a zero-day vulnerability in some SonicWall devices is allowing ransomware attacks.
Ransomware gangs may be exploiting an unknown vulnerability in SonicWall devices to launch attacks on dozens of organizations.
Multiple incident response companies released warnings over the weekend about threat actors using the Akira ransomware to target SonicWall firewall devices for initial access. Experts at Arctic Wolf first revealed the incidents on Friday.
SonicWall has not responded to repeated requests for comment about the breaches but published a blog post on Monday afternoon confirming that it is aware of the campaign.
The company said Arctic Wolf, Google and Huntress have warned over the last 72 hours that there has been an increase in cyber incidents involving Gen 7 SonicWall firewalls that use the secure sockets layer (SSL) protocol.
“We are actively investigating these incidents to determine whether they are connected to a previously disclosed vulnerability or if a new vulnerability may be responsible,” the company said.
SonicWall said it is working with researchers, updating customers and will release updated firmware if a new vulnerability is found.
The company echoed the advice of several security firms, telling customers to disable SonicWall VPN services that use the SSL protocol.
At least 20 incidents
Arctic Wolf said on Friday that it has seen multiple intrusions within a short period of time and all of them involved access through SonicWall SSL VPNs.
“While credential access through brute force, dictionary attacks, and credential stuffing have not yet been definitively ruled out in all cases, available evidence points to the existence of a zero-day vulnerability,” the company said. None of the incident response companies have specified what that bug might be.
“In some instances, fully patched SonicWall devices were affected following credential rotation,” Arctic Wolf said, referring to the process of regularly resetting logins or other access.
The researchers added that the ransomware activity involving SonicWall VPNs began around July 15.
When pressed on whether any recent known SonicWall vulnerabilities are to blame for the attacks, an Arctic Wolf spokesperson said the researchers have “seen fully patched devices affected in this campaign, leading us to believe that this is tied to a net new zero day vulnerability.”
Arctic Wolf said in its advisory that given the high likelihood of such a bug, organizations “should consider disabling the SonicWall SSL VPN service until a patch is made available and deployed.”
Over the weekend, Arctic Wolf’s assessment was backed up by incident responders at Huntress, who confirmed several incidents involving the SonicWall SSL VPN.
A Huntress official said they have seen around 20 attacks since July 25 and many of the incidents include the abuse of privileged accounts, lateral movement, credential theft and ransomware deployment.
“This is happening at a pace that suggests exploitation, possibly a zero day exploit in Sonicwall. Threat actors have gained control of accounts that even have MFA deployed,” the official said.
He confirmed that the incidents Huntress examined also involved Akira ransomware.
'This isn't isolated'
Huntress released a lengthy threat advisory on Monday warning of a “likely zero-day vulnerability in SonicWall VPNs” that was being used to facilitate ransomware attacks. Like Arctic Wolf, they urged customers to disable the VPN service immediately.
“Over the last few days, the Huntress Security Operations Center (SOC) has been responding to a wave of high-severity incidents originating from SonicWall Secure Mobile Access (SMA) and firewall appliances,” Huntress explained.
“This isn't isolated; we're seeing this alongside our peers at Arctic Wolf, Sophos, and other security firms. The speed and success of these attacks, even against environments with MFA enabled, strongly suggest a zero-day vulnerability is being exploited in the wild.”
SonicWall devices are frequent targets for hackers because the types of appliances the company produces serve as gateways for secure remote access.
Just two weeks ago, Google warned of a campaign targeting end-of-life SonicWall SMA 100 series appliances through a bug tracked as CVE-2024-38475.
techcrunch.com - Google’s AI-powered bug hunter has just reported its first batch of security vulnerabilities.
Heather Adkins, Google’s vice president of security, announced Monday that its LLM-based vulnerability researcher Big Sleep found and reported 20 flaws in various popular open source software.
Adkins said that Big Sleep, which is developed by the company’s AI department DeepMind as well as its elite team of hackers Project Zero, reported its first-ever vulnerabilities, mostly in open source software such as audio and video library FFmpeg and image-editing suite ImageMagick.
Given that the vulnerabilities are not fixed yet, we don’t have details of their impact or severity, as Google does not yet want to provide details, which is a standard policy when waiting for bugs to be fixed. But the simple fact that Big Sleep found these vulnerabilities is significant, as it shows these tools are starting to get real results, even if there was a human involved in this case.
“To ensure high quality and actionable reports, we have a human expert in the loop before reporting, but each vulnerability was found and reproduced by the AI agent without human intervention,” Google’s spokesperson Kimberly Samra told TechCrunch.
Royal Hansen, Google’s vice president of engineering, wrote on X that the findings demonstrate “a new frontier in automated vulnerability discovery.”
LLM-powered tools that can look for and find vulnerabilities are already a reality. Other than Big Sleep, there’s RunSybil and XBOW, among others.
Les arnaques avec un faux SMS annonçant une contravention à régler ont fait de nombreuses victimes ces derniers mois. A l’image d’une retraitée, qui a perdu 3000 francs dans l’affaire, comme le relate la «Tribune de Genève» lundi. Le message provenant de sites frauduleux, comme «amendes.ch», «parkings-vd.com» ou «parkings-ge.com», annonce au destinataire qu’il doit s’acquitter de 40 francs et l’invite à cliquer sur un lien. Il s’agit en réalité d’un hameçonnage permettant aux malfrats d’accéder aux données bancaires des victimes.
Plus de 220 personnes ont annoncé avoir fait les frais de cette arnaque à Genève et dans le canton de Vaud, rapporte le quotidien genevois. Le coût total des dommages s’élève à 3,6 millions de francs. Pour éviter la fraude, avait indiqué la police vaudoise notamment, il est recommandé de ne jamais cliquer sur les liens proposés, de bloquer le correspondant et de supprimer le messager.
venturebeat.com - OpenAI abruptly removed a ChatGPT feature that made conversations searchable on Google, sparking privacy concerns and industry-wide scrutiny of AI data handling.
OpenAI made a rare about-face Thursday, abruptly discontinuing a feature that allowed ChatGPT users to make their conversations discoverable through Google and other search engines. The decision came within hours of widespread social media criticism and represents a striking example of how quickly privacy concerns can derail even well-intentioned AI experiments.
The feature, which OpenAI described as a “short-lived experiment,” required users to actively opt in by sharing a chat and then checking a box to make it searchable. Yet the rapid reversal underscores a fundamental challenge facing AI companies: balancing the potential benefits of shared knowledge with the very real risks of unintended data exposure.
How thousands of private ChatGPT conversations became Google search results
The controversy erupted when users discovered they could search Google using the query “site:chatgpt.com/share” to find thousands of strangers’ conversations with the AI assistant. What emerged painted an intimate portrait of how people interact with artificial intelligence — from mundane requests for bathroom renovation advice to deeply personal health questions and professionally sensitive resume rewrites. (Given the personal nature of these conversations, which often contained users’ names, locations, and private circumstances, VentureBeat is not linking to or detailing specific exchanges.)
“Ultimately we think this feature introduced too many opportunities for folks to accidentally share things they didn’t intend to,” OpenAI’s security team explained on X, acknowledging that the guardrails weren’t sufficient to prevent misuse.
scworld.com 04.08 - Aeroflot, Russia's flag carrier, had travel information purportedly from its CEO Sergei Aleksandrovsky leaked by Belarusian hacktivist operation Cyber Partisans after Russian internet watchdog Roskomnadzor refuted any data breach resulting from last week's massive cyberattack that has prompted the cancellation of more than 50 flights, reports The Record, a news site by cybersecurity firm Recorded Future.
Included in the exposed data were information from over 30 flights taken by Aleksandrovsky from April 2024 to June 2025, claimed Cyber Partisans, which threatened the imminent reveal of more stolen data following the theft of Aeroflot's entire flight history database. Cyber Partisans noted that the extensive data compromise was made possible by weak employee credentials and the airline's use of outdated Windows versions. While the legitimacy of the data has not yet been confirmed, it contained Aleksandrovsky's passport number that matched those found in older breaches, according to investigative news outlet The Insider.
tomshardware.com - A leading mobile device insurance and service network has initiated insolvency proceedings in the wake of a cyberattack. Selling properties and cutting staff numbers wasn't enough to save the business.
The Einhaus Group was once a familiar name, with its services available through 5,000 retail outlets in Germany and an annual revenue of around 70 million Euros.
A leading mobile device insurance and service network has initiated insolvency proceedings in the wake of a cyberattack. Germany’s Einhaus Group was targeted by hackers in March 2023 and is understood to have paid a ransom(ware) fee of around $230,000 at the time, according to Wa.de and Golem.de (machine translations). However, the once large and successful company, with partnerships including Cyberport, 1&1, and Deutsche Telekom, struggled to recover from the service interruption and the obvious financial strains, which now appear to be fatal.
The ides of March
In mid-March 2023, Wilhelm Einhaus, founder of the Einhaus Group, recalls coming into the office in the morning to witness a ‘horrific’ greeting. On the output tray of every printer in the office was a page announcing, “We've hacked you. All further information can be found on the dark web.” Further investigations revealed that the hack group 'Royal' was the culprit. They had encrypted all of Einhaus Group’s systems, which were essential for the day-to-day running of the business. 'Royal' demanded a ransom payment, thought to be around $230,000 in Bitcoins, to return access to the computers.
Of course, with operational systems down, there was an immediate impact on Einhaus. The police were involved promptly. However, the affected firm seems to have decided to pay the ransom, as it could see business losses/damages piling up – meaning continuing without the computer systems was untenable. Einhaus estimated that the hacker-inflicted damage to its business was in the mid-seven-figure range.
nltimes.nl - Several major government institutions across the Caribbean part of the Kingdom of the Netherlands were hit by cyberattacks last week, including a ransomware attack on Curaçao’s Tax and Customs Administration that temporarily disabled critical services, NOS reports.
According to Curaçao’s Minister of Finance, ransomware was used in the attack on the tax authority. After the breach was discovered by staff, one of the agency’s systems was taken offline as a precaution. An investigation into the origin and impact of the attack is ongoing. The Ministry of Finance stated that no confidential information was compromised.
Despite the breach, the online platform for filing and paying taxes remained operational. However, both the telephone customer service and in-person assistance were unavailable for several days. All services were restored by Monday, the ministry confirmed.
Meanwhile, the Court of Justice — which operates across all six Caribbean islands of the Kingdom — was also affected by a cyber incident. A virus was detected in the court’s IT system, prompting officials to shut down the entire computer network out of caution. Several court cases scheduled for last week were postponed, although most hearings continued as planned. Restoration efforts are still underway.
In Aruba, hackers also gained unauthorized access to official email accounts belonging to members of parliament. The extent of the breach and potential consequences remain unclear.
In response to the string of incidents, authorities on Sint-Maarten issued a public alert urging businesses and institutions on the islands to increase their cybersecurity vigilance.
The wave of cyberattacks follows a separate hacking incident in the Netherlands just two weeks ago, when the national Public Prosecution Service (Openbaar Ministerie) disconnected all its systems from the internet after detecting a breach. The disruption continues to have major consequences. Defense attorneys have reported significant difficulty accessing essential information, hindering their ability to represent clients.
france24.com - Chinese authorities summoned Nvidia representatives on Thursday to discuss "serious security issues" over some of its artificial intelligence chips, as the US tech giant finds itself entangled in trade tensions between Beijing and Washington.
Nvidia is a world-leading producer of AI semiconductors, but the United States effectively restricts which chips it can export to China on national security grounds.
A key issue has been Chinese access to the "H20", a less powerful version of Nvidia's AI processing units that the company developed specifically for export to China.
The California-based firm said this month it would resume H20 sales to China after Washington pledged to remove licensing curbs that had halted exports.
But the firm still faces obstacles -- US lawmakers have proposed plans to require Nvidia and other manufacturers of advanced AI chips to include built-in location tracking capabilities.
And Beijing's top internet regulator said Thursday it had summoned Nvidia representatives to discuss recently discovered "serious security issues" involving the H20.
The Cyberspace Administration of China said it had asked Nvidia to "explain the security risks of vulnerabilities and backdoors in its H20 chips sold to China and submit relevant supporting materials".
The statement posted on social media noted that, according to US experts, location tracking and remote shutdown technologies for Nvidia chips "are already matured".
The announcement marked the latest complication for Nvidia in selling its advanced products in the key Chinese market, where it is in increasingly fierce competition with homegrown technology firms.
Nvidia committed
CEO Jensen Huang said during a closely watched visit to Beijing this month that his firm remained committed to serving local customers.
Huang said he had been assured during talks with top Chinese officials during the trip that the country was "open and stable".
"They want to know that Nvidia continues to invest here, that we are still doing our best to serve the market here," he said.
Nvidia this month became the first company to hit $4 trillion in market value -- a new milestone in Wall Street's bet that AI will transform the global economy.
Jost Wubbeke of the Sinolytics consultancy told AFP the move by China to summon Nvidia was "not surprising in the sense that targeting individual US companies has become a common tool in the context of US-China tensions".
"What is surprising, however, is the timing," he noted, after the two countries agreed to further talks to extend their trade truce.
"China's action may signal a shift toward a more assertive stance," Wubbeke said.
Beijing is also aiming to reduce reliance on foreign tech by promoting Huawei's domestically developed 910C chip as an alternative to the H20, he added.
"From that perspective, the US decision to allow renewed exports of the H20 to China could be seen as counterproductive, as it might tempt Chinese hyperscalers to revert to the H20, potentially undermining momentum behind the 910C and other domestic alternatives."
New hurdles to Nvidia's operation in China come as the country's economy wavers, beset by a years-long property sector crisis and heightened trade headwinds under US President Donald Trump.
Chinese President Xi Jinping has called for the country to enhance self-reliance in certain areas deemed vital for national security -- including AI and semiconductors -- as tensions with Washington mount.
The country's firms have made great strides in recent years, with Huang praising their "super-fast" innovation during his visit to Beijing this month.
therecord.media 04.08 - Researchers have discovered more than 10 patents for powerful offensive cybersecurity technologies filed by a prominent Chinese company allegedly involved in Beijing’s Silk Typhoon campaign.
Researchers have discovered more than 10 patents for powerful offensive cybersecurity technologies filed by a prominent Chinese company allegedly involved in Beijing’s Silk Typhoon campaign.
SentinelOne's threat researchers pored through recent Justice Department indictments of prominent Chinese hackers and mapped out the country’s evolving web of private companies that are hired to launch cyberattacks on behalf of the government.
The report focuses on intellectual property rights filings by Shanghai Firetech, a company the DOJ said works on behalf of the Shanghai State Security Bureau (SSSB). The company was allegedly involved in many of the Silk Typhoon attacks and was previously identified as part of the Hafnium attacks seen in 2021.
The researchers found previously unseen patents on offensive technologies tied to Shanghai Firetech, SentinelLabs expert Dakota Cary told Recorded Future News.
The findings suggest the company “serves other offensive missions not tied to the Hafnium cluster,” he said.
“The company also has patents on a variety of offensive tools that suggest the capability to monitor individual's homes, like ‘intelligent home appliances analysis platform,’ ‘long-range household computer network intelligentized control software,’ and ‘intelligent home appliances evidence collection software’ which could support surveillance of individuals abroad.”
Cary noted that intelligence agencies like the CIA are known to use similar tools.
Shanghai Firetech also filed patents for software for “remote” evidence collection, and for targeting routers and Apple devices, among other uses.
The patent for Apple computers stood out to the researchers because it allows actors to remotely recover files from devices and was not previously documented as a capability of any Hafnium-related threat actor.
SentinelLabs said the technologies “offer strong, often previously unreported offensive capabilities, from acquisition of encrypted endpoint data, mobile forensics, to collecting traffic from network devices.”
The Justice Department indicted two prominent hackers this month — Xu Zewei and Zhang Yu — that are accused of working with China’s Ministry of State Security (MSS) and its Shanghai bureau. The indictments said Xu and Zhang worked for two firms previously unattributed in the public domain to the Hafnium/Silk Typhoon group.
Xu was arrested after flying into Milan on July 3, and prosecutors accused both men of being deeply involved in China’s cyberattacks on institutions working on COVID-19 vaccines throughout 2020 and 2021. The DOJ obtained emails from Xu to the Shanghai security bureau confirming he had acquired the contents of the COVID-19 researchers’ mailboxes.
nytimes.com 04.08 - The introduction of a state-approved messaging app has raised fears that Russia could be preparing to block WhatsApp and Telegram.
Russia is escalating its efforts to curtail online freedom, taking new steps toward a draconian state-controlled internet.
The authorities are cracking down on workarounds that Russians have been using for access to foreign apps and banned content, including through new laws signed by President Vladimir V. Putin this past week. Moscow has also been impeding the function of services from U.S. tech companies, like YouTube, that Russians have used for years.
At the same time, the Kremlin is building out a domestic ecosystem of easily monitored and censored Russian alternatives to Western tech products. That includes a new state-approved messaging service, MAX, which will come preinstalled by law on all new smartphones sold in Russia starting next month.
The idea, experts say, is to migrate more Russians from an open internet dominated by the products of Western tech giants to a censored online ecosystem, where Russians primarily use software under the gaze and influence of the state. The effort has advanced significantly amid wartime repression, but it is unclear how far it will go.
“The goal here is absolute control,” said Anastasiia Kruope, a researcher at Human Rights Watch who wrote a recent report on declining Russian internet freedoms.
The Kremlin wants to control not only the information available online but also where and how internet traffic flows, Ms. Kruope said, so the Russian internet can function in isolation and be switched on and off at will. Russia’s technical capabilities for clamping down are improving, she added.
“They are not perfect,” Ms. Kruope said. “They are not nearly at the level they would like them to be. But they are getting better, and this is the reason to start paying attention.”“The goal here is absolute control,” said Anastasiia Kruope, a researcher at Human Rights Watch who wrote a recent report on declining Russian internet freedoms.
The Kremlin wants to control not only the information available online but also where and how internet traffic flows, Ms. Kruope said, so the Russian internet can function in isolation and be switched on and off at will. Russia’s technical capabilities for clamping down are improving, she added.
“They are not perfect,” Ms. Kruope said. “They are not nearly at the level they would like them to be. But they are getting better, and this is the reason to start paying attention.”
blog.pypi.org - - The Python Package Index Blog - PyPI Users are receiving emails detailing them to log in to a fake PyPI site.
PyPI has not been hacked, but users are being targeted by a phishing attack that attempts to trick them into logging in to a fake PyPI site.
Over the past few days, users who have published projects on PyPI with their email in package metadata may have received an email titled:
[PyPI] Email verification
from the email address noreply@pypj.org.
Note the lowercase j in the domain name, which is not the official PyPI domain, pypi.org.
This is not a security breach of PyPI itself, but rather a phishing attempt that exploits the trust users have in PyPI.
The email instructs users to follow a link to verify their email address, which leads to a phishing site that looks like PyPI but is not the official site.
The user is prompted to log in, and the requests are passed back to PyPI, which may lead to the user believing they have logged in to PyPI, but in reality, they have provided their credentials to the phishing site.
PyPI Admins are looking into a few methods of handling this attack, and want to make sure users are aware of the phishing attempt while we investigate different options.
There is currently a banner on the PyPI homepage to warn users about this phishing attempt.
Always inspect the URL in the browser before logging in.
We are also waiting for CDN providers and name registrars to respond to the trademark and abuse notifications we have sent them regarding the phishing site.
If you have received this email, do not click on any links or provide any information. Instead, delete the email immediately.
If you have already clicked on the link and provided your credentials, we recommend changing your password on PyPI immediately. Inspect your account's Security History for anything unexpected.
reuters.com - July 30 (Reuters) - More than 90 state and local governments have been targeted using the recently revealed vulnerability in Microsoft server software, according to a U.S. group devoted to helping local authorities collaborate against hacking threats.
The nonprofit Center for Internet Security, which houses an information-sharing group for state, local, tribal, and territorial government entities, provided no further details about the targets, but said it did not have evidence that the hackers had broken through.
None have resulted in confirmed security incidents," Randy Rose, the center's vice president of security operations and intelligence, said in an email.
A wave of hacks hit servers running vulnerable versions of Microsoft SharePoint this month, causing widespread concern. The campaign has claimed at least 400 victims, according to Netherlands-based cybersecurity firm Eye Security. Multiple federal government agencies are reportedly among the victims, and new ones are being identified every day.
On Wednesday, a spokesperson for one of the U.S. Department of Energy's 17 national labs said it was among those hit.
"Attackers did attempt to access Fermilab's SharePoint servers," the spokesperson said, referring to the U.S. Fermi National Accelerator Laboratory. "The attackers were quickly identified, and the impact was minimal, with no sensitive or classified data accessed." The Fermilab incident was first reported by Bloomberg.
The U.S. Department of Energy has previously said the SharePoint security hack has affected "a very small number" of its systems
channelnewsasia.com - The decision to identify cyber threat group UNC3886 was because Singaporeans “ought to know about it” given the seriousness of the threat, said the minister.
SINGAPORE: While naming a specific country linked to cyber threat group UNC3886 is not in Singapore’s interest at this point in time, the attack was still serious enough for the government to let the public know about the group, said Coordinating Minister for National Security and Minister for Home Affairs K Shanmugam on Friday (Aug 1).
Speaking to reporters on the side of the Cyber Security Agency of Singapore’s (CSA) Exercise Cyber Star, the national cybersecurity crisis management exercise, Mr Shanmugam said that when it comes to naming any country responsible for a cyber attack, “we always think about it very carefully”.
Responding to a question from CNA on reports tying the group to China, Mr Shanmugam said: “Media coverage (and) industry experts all attribute UNC3886 to some country … Government does not comment on this.
“We release information that we assess is in the public interest. Naming a specific country is not in our interest at this point in time.”
UNC3886 has been described by Google-owned cybersecurity firm Mandiant as a "China-nexus espionage group" that has targeted prominent strategic organisations on a global scale.
Mr Shanmugam had announced on Jul 18 that Singapore is actively dealing with a "highly sophisticated threat actor" that is attacking critical infrastructure, identifying the entity as UNC3886 without disclosing if it was a state-linked actor.
He said the threat actor poses a serious danger to Singapore and could undermine the country's national security, and added that it was not in Singapore's security interests to disclose further details of the attack then.
When asked the following day about UNC3886's alleged links to China and possible retaliation for naming them, Mr Shanmugam, who is also Home Affairs Minister, said this was "speculative".
"Who they are linked to and how they operate is not something I want to go into," he said.
Responding to media reports in a Jul 19 Facebook post, the Chinese embassy in Singapore expressed its "strong dissatisfaction" at the claims linking the country to UNC3886, stating that they were "groundless smears and accusations against China".
“In fact, China is a major victim of cyberattacks," it wrote.
"The embassy would like to reiterate that China is firmly against and cracks down (on) all forms of cyberattacks in accordance with law. China does not encourage, support or condone hacking activities."
On Friday, Mr Shanmugam also gave his reasons for disclosing the identity of threat actors like UNC3886.
“We look at the facts of each case (and) the degree of confidence we have before we can name. And when we decide to name the threat actor, we look at whether it is in Singapore's best interest,” said Mr Shanmugam, who is also the home affairs minister.
In this case, the threat, attack and compromise to Singapore’s infrastructure was “serious enough” and the government was confident enough to name UNC3886 as the perpetrators, he said.
“Here, we said this is serious. They have gotten in. They are compromising a very serious critical infrastructure. Singaporeans ought to know about it, and awareness has got to increase. And because of the seriousness, it is in the public interest for us to disclose,” said Mr Shanmugam.
therecord.media (01.08.2025) - Authorities in Luxembourg said a nationwide telecommunications outage in July was caused by a deliberately disruptive cyberattack. Huawei networking products were reportedly the target.
Luxembourg’s government announced on Thursday it was formally investigating a nationwide telecommunications outage caused last week by a cyberattack reportedly targeting Huawei equipment inside its national telecoms infrastructure.
The outage on July 23 left the country’s 4G and 5G mobile networks unavailable for more than three hours. Officials are concerned that large parts of the population were unable to call the emergency services as the fallback 2G system became overloaded. Internet access and electronic banking services were also inaccessible.
According to government statements issued to the country’s parliament, the attack was intentionally disruptive rather than an attempt to compromise the telecoms network that accidentally led to a system failure.
Officials said the attackers exploited a vulnerability in a “standardised software component” used by POST Luxembourg, the state-owned enterprise that operates most of the country’s telecommunications infrastructure. The government’s national alert system, which officials had intended to use to warn the population about the incident, failed to reach many people because it also depends on POST’s mobile network.
POST’s director-general described the attack itself as “exceptionally advanced and sophisticated,” but stressed it did not compromise or access internal systems and data. POST itself and the national CSIRT are currently forensically investigating the cause of the outage.
Although the government’s statements avoid naming the affected supplier, Luxembourg magazine Paperjam reported the attack targeted software used in Huawei routers. Paperjam added that the country’s critical infrastructure regulator is currently asking any organisations using Huawei enterprise routers to contact the CSIRT.
Remote denial-of-service vulnerabilities have previously been identified in the VRP network operating system used in Huawei’s enterprise networking products, although none have recently been publicly identified. Huawei’s press office did not respond to a request for comment.
The Luxembourg government convened a special crisis cell within the High Commission for National Protection (HCPN) to handle the response to the incident and to investigate its causes and impacts, alongside the CSIRT and public prosecutor.
The CSIRT’s full forensic investigation is intended to confirm how the attack happened, while the public prosecutor will assess whether a crime has taken place and if a perpetrator can be identified and prosecuted.
The incident has also accelerated Luxembourg’s national resilience review, a process already underway before the attack. Authorities, concerned that a single point of failure had such a dramatic disruptive effect, are now reassessing the robustness of critical infrastructure, including fallback procedures for telecom and emergency services.
Luxembourg is also exploring regulatory changes to allow mobile phones to automatically switch to other operators’ networks during telecom outages, a practice already used in countries like the United Kingdom, Germany and the United States for emergency calls.
techcrunch.com 24.07 - "We're getting a lot of stuff that looks like gold, but it's actually just crap,” said the founder of one security testing firm. AI-generated security vulnerability reports are already having an effect on bug hunting, for better and worse.
So-called AI slop, meaning LLM-generated low-quality images, videos, and text, has taken over the internet in the last couple of years, polluting websites, social media platforms, at least one newspaper, and even real-world events.
The world of cybersecurity is not immune to this problem, either. In the last year, people across the cybersecurity industry have raised concerns about AI slop bug bounty reports, meaning reports that claim to have found vulnerabilities that do not actually exist, because they were created with a large language model that simply made up the vulnerability, and then packaged it into a professional-looking writeup.
“People are receiving reports that sound reasonable, they look technically correct. And then you end up digging into them, trying to figure out, ‘oh no, where is this vulnerability?’,” Vlad Ionescu, the co-founder and CTO of RunSybil, a startup that develops AI-powered bug hunters, told TechCrunch.
“It turns out it was just a hallucination all along. The technical details were just made up by the LLM,” said Ionescu.
Ionescu, who used to work at Meta’s red team tasked with hacking the company from the inside, explained that one of the issues is that LLMs are designed to be helpful and give positive responses. “If you ask it for a report, it’s going to give you a report. And then people will copy and paste these into the bug bounty platforms and overwhelm the platforms themselves, overwhelm the customers, and you get into this frustrating situation,” said Ionescu.
“That’s the problem people are running into, is we’re getting a lot of stuff that looks like gold, but it’s actually just crap,” said Ionescu.
Just in the last year, there have been real-world examples of this. Harry Sintonen, a security researcher, revealed that the open source security project Curl received a fake report. “The attacker miscalculated badly,” Sintonen wrote in a post on Mastodon. “Curl can smell AI slop from miles away.”
In response to Sintonen’s post, Benjamin Piouffle of Open Collective, a tech platform for nonprofits, said that they have the same problem: that their inbox is “flooded with AI garbage.”
One open source developer, who maintains the CycloneDX project on GitHub, pulled their bug bounty down entirely earlier this year after receiving “almost entirely AI slop reports.”
The leading bug bounty platforms, which essentially work as intermediaries between bug bounty hackers and companies who are willing to pay and reward them for finding flaws in their products and software, are also seeing a spike in AI-generated reports, TechCrunch has learned.
