Also: Norway’s Equinor Hacked, Project Zero Cuts Deadline Disclosure Timeline
Anviksha More (AnvikshaMore) •
November 19, 2026
Image: Shutterstock/ISMG
Every week, ISMG rounds up cybersecurity incidents and breaches around the world. This week: European Commission hacked, Google’s Project Zero cut its disclosure deadline to 30 days, Equinor hacked, an Osaka data breach, Ruby2Shell debuts an Irish credit union mass breach. 8-Digits Team ransomware suspect extradited to U.S., ransomware attacks rise but payments fall, Ukraine and FBI arrest four in $40 million extortion scheme, Premise Health insider attack now affects 2.8 million.See Also: Why Banks Are Rethinking Fraud and Cybersecurity Together
European Commission Confirms AI-Driven Cyberattack
The European Commission said Thursday that it was targeted in an artificial intelligence-driven cyberattack and is investigating whether any data was breached.
The incident stemming from the “misuse of an AI tool” was detected on Oct. 22 and confirmed following an initial investigation on Oct. 26. Several internal human resources applications, including time management and medical reimbursement systems, were temporarily taken offline out of preventative measures, but most have been restored.
“We can confirm that this was an AI-generated cyber campaign,” Thomas Regnier, spokesperson for the executive branch of the European Union, said in a press briefing Thursday in Brussels. “At this stage of the investigation, we have not detected any evidence of a data breach affecting commission staff as a whole.”
Regnier said the commission did not receive any ransomware request and has not identified the actor behind the attack.
The intrusion followed warnings about hackers using AI penetration testing framework Artex to target the European Union. “CERT-EU has observed similar instances of AI-driven activities against its constituency. These activities highly likely involved the use of ARTEX,” the European cybersecurity service said this month.
Google Project Zero Cuts Deadline to 30 Days
Google’s vulnerability hunting team Project Zero will shorten the time software vendors have to fix reported security flaws before their details become public. The team cited the growing speed of artificial intelligence-assisted vulnerability discovery and exploitation.
Project Zero is replacing its existing “90+30” disclosure policy with a fixed deadline that will eventually require vendors to address flaws within 30 days. The transition began Nov. 13, when Google removed the additional 30-day grace period previously allowed for patch deployment.
Under the revised policy, Project Zero will publicly disclose vulnerabilities 90 days after reporting them to vendors. The deadline will fall to 60 days on Feb. 2, 2027, and to 30 days on Jan. 3, 2028. Disclosure will take place regardless of whether vendors have released a fix.
The team said large language models have accelerated vulnerability discovery and exploitation, increasing the likelihood that attackers will independently identify the same flaws. AI tools also make it easier for attackers to reverse-engineer patches and develop exploits once security updates become available.
“Patch soak time is no longer meaningful in a world where attackers can almost immediately determine bug details from a patch,” Project Zero security engineer Natalie Silvanovich wrote in the policy announcement.
Google said it consulted vendors and open-source maintainers before introducing the changes. Some raised concerns that an immediate shift to 30-day deadlines could expose users to unpatched vulnerabilities. Vendors also warned that aggressive deadlines could cause them to prioritize newly reported flaws over potentially more serious issues already awaiting remediation.
The phased approach is intended to give vendors time to improve their vulnerability management and patching processes.
The revised policy retains limited exceptions. Vulnerabilities actively exploited in the wild will follow Google’s separate seven-day disclosure policy. Project Zero may also disclose flaws earlier if their details have already become public or if the affected vendor requests disclosure.
Google said it will continue tracking whether vulnerabilities have been fixed, but patch availability will no longer determine when technical details become public.
Equinor Detects Cyberattack on IT Systems
Norwegian energy company Equinor detected an ongoing cyberattack targeting some of its IT systems, prompting the company to mobilize security resources and notify national authorities. The incident has not affected oil and gas production, and the company has found no evidence of data theft.
Equinor, Europe’s largest gas supplier, said Tuesday that it was working to neutralize the attack and restore the integrity of affected systems.
Norwegian news agency NTB reported that the attack had disrupted the company’s access to IT applications and internal systems. An employee familiar with the incident told Reuters, on condition of anonymity, that Equinor’s internal email system was also down.
The Norwegian National Security Authority said it was aware of the incident and was providing assistance as needed.
Osaka Data Breach Exposes 1 Million Email Addresses
The city of Osaka, Japan, warned that more than 1 million email addresses were exposed in a data breach involving a third-party technology provider used to operate its online administrative services platform.
According to the city government, attackers accessed the Arara email distribution service, which is used by TKC Corporation, the contractor responsible for maintaining the platform. The exposed information includes email addresses of people who received messages from the city’s system between July and November.
The incident is linked to a wider breach at Arara, which disclosed on Nov. 2 that a cyberattack had exposed 77 million email addresses belonging to customers across nearly 400 organizations. Those affected include TKC and gaming company Sega.
The Osaka incident follows a separate breach disclosed by the Osaka Prefecture government involving information belonging to participants in a food assistance program.
Japan has faced a growing wave of cyberattacks targeting businesses and public institutions. The government recently announced emergency cybersecurity measures, including increased use of artificial intelligence for cyber defense and about $63 million in emergency funding to strengthen its response capabilities (see: Sony ISP Breach Exposes 2.6 Million User Records).
Don’t Panic Over Ruby2Shell
There’s another new remote code execution vulnerability, this time in the Ruby programing language. Ruby2Shell made its debut in the footsteps of PHP2shell) and Nginx2root. The vulnerability is “a critical one-byte heap overflow in how Ruby changes the case of strings in some legacy single-byte encodings,” Aikido Security explains.
Ruby maintainers disclosed the flaw and its patches on Wednesday. So far no exploitation is apparent, although the maintainers warned that working proof of concepts do exist.
“It’s barely been a day, and we have yet another gem with a 2Shell name,” noted Yordan Ganchev, principal threat intelligence specialist cybersecurity firm watchtower. But system administrators should stay calm. “Finding a vulnerable sink to trigger the vulnerability is trivial, but fully turning it into a weaponized, universal exploit that can land a shell would be challenging, to say the least. A lot of planets and stars would need to align in just the right order for an attacker to correctly apply heap feng shui to shape the memory layout,” he said.
Data Breach Hits 40 Irish Credit Unions Through IT Provider
A cyberattack on Belfast-based IT provider Wellington IT exposed sensitive payment information belonging to members of approximately 40 Irish credit unions.
Wellington IT supplies technology services to about 50 credit unions, including 43 in Ireland. The attacker exploited an authorization flaw in the company’s infrastructure.
The Irish Examiner reported that an automated script made approximately 5 million requests to exploit the vulnerability. The attacker reportedly accessed private records using valid credentials.
Exposed information potentially includes payee names, international bank account numbers, bank identifier codes, account numbers, sort codes, addresses and payment references saved in members’ online banking platforms.
The number of affected individuals remains unknown. Wellington IT reportedly told customers that it could not determine exactly which records were accessed because it lacked the necessary activity logs.
The company said members’ account passwords, PINs, payment card details and online banking credentials were not affected. However, credit unions have advised members to watch for suspicious communications and impersonation attempts.
8-Digits Team Ransomware Suspect Extradited to U.S.
Italy extradited a 43-year-old Kazakh national to the United States to face charges of participating in the 8-Digits Team ransomware group, which targeted U.S. hospitals, healthcare providers and schools in a campaign that generated more than $500 million in ransom demands.
Roman Khlynovskiy made his initial appearance in federal court in Knoxville, Tennessee, on Nov. 16. He is accused of participating in a conspiracy affiliated with the INC and Lynx ransomware operations.
According to the indictment, the group targeted victims in at least six U.S. states, infiltrated their networks and stole sensitive information before demanding ransom payments to prevent its publication.
The attackers claimed to possess more than 50 terabytes of stolen data, including over 70 million confidential health records. The campaign generated more than $500 million in ransom demands, the DOJ said.
Khlynovskiy faces charges including conspiracy to commit computer fraud and identity theft, wire fraud, extortion involving computers and money laundering. He is also charged under the Take It Down Act for threats to disclose nonconsensual intimate images.
If convicted on all charges, he faces a maximum prison sentence of 27 years.
Ransomware Attacks Rise as AI Adoption Grows, Payments Fall
Ransomware attacks increased nearly 40% in the third quarter of 2026 as cybercriminals turned to artificial intelligence to improve phishing campaigns and identify sensitive data, even as fewer victims agreed to pay ransom demands.
Ransomware groups claimed 2,766 breaches during the quarter, up from the previous quarter, according to threat intelligence data from cybersecurity firm Halcyon. The number of groups claiming successful attacks also increased from 75 in the first quarter to 111 in the third quarter.
AI tools, particularly large language models, are helping attackers create convincing phishing messages in multiple languages and identify valuable information in stolen files. This has lowered the barriers for smaller criminal groups to conduct attacks.
Increased attack activity has not translated into higher ransom payments. Sophos found that only 32% of retailers paid ransom demands in 2026, compared with 58% in 2025. About 71% recovered their data from backups, with most restoring operations within a week.
The median ransom demand also declined to $700,000 in 2026 from $2 million in 2024, according to Sophos.
Ukraine, FBI Arrest 4 Ransomware Suspects in $40M Extortion Scam
Ukrainian authorities, working with the FBI, arrested four suspected members of an international ransomware organization linked to more than $40.5 million in ransom payments from at least 28 companies worldwide.
The arrests follow a two-year investigation into a criminal organization comprising at least 27 individuals from Ukraine, Kazakhstan, Russia and other countries. Investigators say the network was connected to a wider cybercrime operation involving more than 60 members across eight jurisdictions, including former members of the Conti ransomware gang.
The attackers targeted organizations in the oil and gas, pharmaceutical, manufacturing, logistics and financial sectors, as well as critical infrastructure.
According to Ukraine’s National Police, the group impersonated IT support personnel and used an artificial intelligence-powered platform to generate voice messages for fraudulent calls. Attackers persuaded employees to disclose login credentials or install remote-access software, enabling them to infiltrate corporate networks, steal data and encrypt systems.
In one incident, the group stole 1.4 terabytes of data from a U.S. company and initially demanded $10 million, eventually receiving $5 million after negotiations.
Ukrainian police conducted 66 simultaneous searches across Kyiv and 11 regions, seizing computers, mobile phones, bank cards, nearly $370,000 and 20,000 euros in cash.
The four suspects face charges under Ukrainian law and could receive prison sentences of up to 15 years if convicted. The investigation is ongoing.
Premise Health Insider Breach Affects 2.76 Million
An insider data breach at Premise Health Holding, an employer-sponsored healthcare organization, has affected more than 2.76 million individuals across the United States, according to the company’s breach report submitted to Oregon state regulators.
The incident appears to be on track to rank among the largest insider health data breaches reported in the U.S. to date.
In recent weeks, several state attorneys general had publicly posted breach reports submitted by Premise, including in Texas, South Carolina and Washington state, but those filings only disclosed the number of people affected in those states (see: Premise Health Insider Breach Could Affect Millions).
The number of people affected nationwide was not made public until the Oregon report by Premise was posted this week.
The Tennessee company, which provides services to 11 million members in 46 states, said in a breach notice that an employee was discovered on Sept. 18 of violating Premise’s data security policies and misappropriating certain information, for about six months, beginning in March.
Premise Health said it fired the employee, revoked the individual’s credentials and system access, validated its security controls, launched an investigation and reported the incident to law enforcement.
The compromised information varied by individual but potentially included names, addresses, dates of birth, government identification numbers such as Social Security numbers, health insurance information, diagnostic and treatment information, and other data provided in connection with their eligibility for services or their relationship with Premise Health.
Other Stories From This Week
With reporting by ISMG’s Tiffany Wang in Manhattan, Marianne Kolbasuk McGee in the Boston exurbs and David Perera in Northern Virginia.
Click Here For The Original Source.
——————————————————–
