AI tools from Google and OpenAI helped cybersecurity firm Wiz discover critical vulnerabilities in PHP, a programming language that runs on over 70% of websites worldwide. The flaws, known as PHP2Shell, could allow attackers to hack into websites and access sensitive data like passwords. Governments across Europe, including the Netherlands, France, Denmark, and Finland, took sites offline as a precaution, while major Japanese companies such as Sega, All Nippon Airways, and MUFG Bank also suspended online services.
The US National Security Agency and Europol warned organizations about the risks on Wednesday. Researchers compared the severity of the vulnerabilities to the 2014 Heartbleed bug, one of the most notable security incidents in internet history. OpenAI described the discovery as one of the most impactful security issues in recent memory, and developers rushed to release patches for the long-hidden flaws.
In other AI-related news, a new phishing-as-a-service platform called Athena StealthRoute is using AI to generate malicious campaigns. Abnormal Security detected 89,700 attack messages targeting 2,280 organizations in the United States and Europe during an August 2026 campaign. The platform includes an AI Studio with multiple workflows for creating phishing lures and methods that make traditional link-based detection less effective.
Research from Harvard and MIT found that AI models give different salary negotiation advice based on gender. Studying 86,000 AI responses, the researchers found all four models recommended higher salary targets for men, while women were more often advised to request flexible hours or parental leave. The study suggests AI systems may reinforce traditional gender roles rather than reduce inequalities.
Mexico is investing in its tech sector through new semiconductor initiatives and an AI training center, with plans to certify 25,000 professionals in AI and data science. Meanwhile, Target is rolling out an AI shopping assistant ahead of the holiday season to help customers find products more easily. Kent RO Systems is using AI across its product lines, including a voice assistant called Kent Ojas that works in nine Indian languages.
Key Takeaways
- AI tools from Google and OpenAI helped discover critical PHP vulnerabilities affecting over 70% of websites worldwide.
- The PHP2Shell bugs could let attackers access sensitive information like passwords on compromised websites.
- Governments in the Netherlands, France, Denmark, and Finland took sites offline as a precaution.
- Major Japanese companies including Sega, All Nippon Airways, and MUFG Bank suspended online services due to the flaws.
- The US NSA and Europol issued warnings about the vulnerabilities, compared in severity to the 2014 Heartbleed bug.
- A new AI phishing platform called Athena StealthRoute targeted 2,280 organizations with 89,700 attack messages in August 2026.
- Harvard and MIT research found AI models recommend higher salaries for men while steering women toward flexible work requests.
- Mexico plans to certify 25,000 professionals in AI and data science through new training programs.
- Target is launching an AI shopping assistant in time for the holiday sales season.
- Kent RO Systems uses AI across its products, including a voice assistant supporting nine Indian languages.
AI finds critical flaw sending governments scrambling
Government websites in Europe and elsewhere went offline after critical vulnerabilities were found in PHP, a programming language used on over 70% of websites. Developers released urgent fixes after the flaws were discovered using AI tools. Cyber company Wiz found the flaw using tools from Google and OpenAI. Countries including the Netherlands, France, Denmark, and Finland took sites offline as a precaution.
AI reveals massive internet security risk in PHP
Governments and cyber-security agencies worldwide are scrambling to contain a massive flaw in PHP, which powers over 70% of websites including Wikipedia and shopping sites. The cybersecurity firm Wiz called it the most severe issue ever detected. OpenAI, whose technology helped find the flaws, called it one of the most impactful security issues in recent memory. Developers rolled out patches to fix the long-hidden vulnerabilities.
Government websites go dark after AI flaw found
Government websites around the world went offline after critical vulnerabilities were disclosed in PHP, a widely used programming language. Developers released urgent fixes and encouraged all users to update immediately. Tech firms warned that some hackers were already exploiting the flaws. Government agencies across Europe took sites offline as a precaution.
AI finds PHP flaws disrupting websites across Asia
Major Japanese companies including Sega, All Nippon Airways, and MUFG Bank suspended websites and online services due to critical flaws in PHP. Authorities across Asia called for urgent security updates. PHP runs on nearly 70% of all websites whose programming language can be identified. Two key vulnerabilities were found in the widely used programming language.
Websites shut down globally after AI finds PHP flaw
Government and company websites worldwide went dark as operators scrambled to patch systems after AI helped find critical vulnerabilities in PHP. The bugs, called PHP2Shell, could let attackers hack into websites and access sensitive information like passwords. The US National Security Agency and Europol warned organizations on Wednesday. Companies like ANA, JR West, MUFG Bank, and Wikipedia took precautionary action, while researchers compared the severity to the 2014 Heartbleed bug.
Mexico boosts semiconductor and AI training programs
Mexico is strengthening its tech industry through new semiconductor initiatives and an AI training center. The country plans to certify 25,000 professionals in AI and data science. The government is investing in these programs to create jobs and attract foreign investment. Mexico showcased its semiconductor capabilities at SEMICON West 2026.
Where AI helps and hurts in real estate listing marketing
AI can help real estate agents save time on captions and photo cleanup, but it struggles with exact facts and brand consistency. Generated flyers can get details wrong, like listing three bathrooms instead of two, and there is no easy way to fix them. The best approach keeps important data and brand assets locked in templates. AI handles the loose parts while templates fill in the exact details from listing data.
Baltimore County school board candidates debate AI in classrooms
Baltimore County school board candidates shared different views on whether AI should be used in classrooms. Some candidates like Candice Logan-Washington want students to learn to use AI responsibly without it replacing traditional learning. Others like Lisa Mack believe AI does not belong in schools and students should learn old-fashioned research skills. A new state law requires the district to put guardrails around AI technology use.
New AI phishing platform targets thousands of organizations
Athena StealthRoute is a new phishing-as-a-service platform that uses artificial intelligence to create malicious campaigns. Researchers at Abnormal Security found it during an August 2026 campaign that detected 89,700 attack messages targeting 2,280 organizations in the United States and Europe. The platform includes an AI Studio with six workflows for generating phishing lures and tools to hide malicious links. Experts warn that its cloaking methods make traditional link-based detection less effective.
AI gives men and women different salary negotiation advice
A study by Harvard and MIT found that AI gives different salary negotiation advice based on gender. Researchers studied 86,000 AI responses and found all four AI models recommended higher salary targets for men. Women were more often advised to ask for flexible hours and parental leave instead. The findings suggest AI systems may reinforce traditional gender roles rather than reduce inequalities.
Jenna Ortega says artificial intelligence lacks creativity
Jenna Ortega stars in the new film Klara and the Sun, based on a novel by Kazuo Ishiguro. The movie comes out Oct. 23 and is directed by Taika Waititi. Ortega said AI is not creative and cannot replace humans. She prefers audiences see the film like a sci-fi movie such as The Terminator.
AI model uses facial photos to predict cancer survival
Researchers at Mass General Brigham built AI models using 40 million facial images. The models may predict cancer survival rates and estimate biological age. The researchers say these tools are not yet ready for clinical use. NBC News reporters explained how the technology works.
Target launches AI shopping assistant for holidays
Target is rolling out an AI shopping assistant in time for the holiday sales season. The tool is part of the company's push to improve the shopping experience. It aims to help customers find products more easily during the busy holiday period.
Kent RO uses AI to transform products and operations
Kent RO Systems is using AI and IoT technology across its water purifiers, kitchen appliances, and energy products. Connected products generate data that helps the company improve development and customer service. Kent's AI voice assistant called Kent Ojas works in nine Indian languages. The company began exploring AI in 2021 and uses robotic process automation for data tasks.
Sources
- Government websites go dark after AI-discovered flaw sparks security scramble
- How did AI discover the internet's massive cyber-security risk?
- Government websites go dark after AI-discovered flaw sparks security scramble
- AI-detected PHP security flaws disrupt websites, spark warnings across Asia
- Some websites shut down globally as operators address PHP issue after AI finds critical flaw
- Mexico Advances Semiconductor Supply Chain and AI Talent Programs
- Where AI belongs in your listing marketing, and where it doesn’t
- Should AI be used in Baltimore County classrooms? Here’s what school board candidates think.
- New phishing-as-a-service platform uses artificial intelligence
- Study finds AI gives men and women different salary negotiation advice
- Jenna Ortega of 'Klara and the Sun' slams artificial intelligence
- AI model uses facial photos to estimate cancer survival and biological age
- Target rolls out AI shopping assistant as part of holiday push
- From connected products to intelligent operations: How Kent RO is reimagining its business with AI
Comments
Please log in to post a comment.