93 stories tagged Security · page 2 of 4
The Trump administration reportedly asked OpenAI to stagger the release of GPT-5.6 due to security concerns, leading CEO Sam Altman to tell employees the company would launch a limited preview for a small group of users before a wider rollout.
Researchers have demonstrated that AI coding assistants can be tricked into executing malware from GitHub repositories that appear clean. This new attack vector exploits how AI agents automatically process and run code, bypassing human review.
Anthropic claims Alibaba attempted to access its AI model, Claude, through deceptive means. This highlights growing tensions in the competitive AI industry.
OpenAI has launched a new initiative to help find and patch bugs in open-source software, using AI to scan code for vulnerabilities and suggest fixes. This could make widely used tools more secure for everyone.
A developer discovered a serious security flaw in his app months after launch. This highlights risks in the trend of quickly building apps with AI tools. Here's how to protect yourself.
OpenAI's new initiative, Patch the Planet, uses AI and expert review to help open-source maintainers find and fix vulnerabilities. This could make open-source software safer for everyone.
Microsoft has fixed two zero-day vulnerabilities that were publicly disclosed by a security researcher known as Nightmare Eclipse. The patches come amid a heated rivalry between the researcher and Microsoft over disclosure practices.
The US government forced Anthropic to pull its latest AI models, Fable 5 and Mythos 5, over national security concerns. Critics argue the move may be backfiring, boosting Anthropic's reputation by highlighting the models' advanced capabilities.
Microsoft researchers discovered a vulnerability called AutoJack that lets attackers execute arbitrary code on the host system running certain AI agents when a user simply visits a malicious web page.
The US government blocked the release of Anthropic’s latest AI models, Fable 5 and Mythos 5, citing national security concerns. Cybersecurity experts have signed an open letter arguing the restrictions are misguided and dangerous, while Anthropic itself notes that similar vulnerabilities exist in other AI systems available on the market.
Scientists have developed a framework to control AI agents that can act independently. This could help prevent security and privacy risks as these systems become more common.
Cloudflare has launched a new feature for its security platform that allows AI agents to obtain temporary, short-lived credentials. This enables secure, automated access to online services without sharing permanent passwords, reducing attack surface and improving convenience.
Researchers introduced OSGuard, a new benchmark to test if AI agents complete tasks safely. It checks for risky shortcuts that might bypass security or ethics rules.
Amazon CEO Andy Jassy reportedly raised security concerns about Anthropic's AI models, leading to a government crackdown and the models being restricted worldwide. This highlights the growing tensions between tech companies and regulators over AI safety.
Google is suing a Chinese cybercrime group for using AI to send millions of fraudulent text messages. The operation targeted hundreds of thousands of victims, highlighting the growing use of AI in cybercrime.
OpenAI's June 2026 report highlights emerging threats from AI misuse, including deepfakes and automated hacking. The report also outlines steps to mitigate these risks and improve AI safety.
Researchers discovered a new tracking method that analyzes SSD activity through JavaScript. This could let websites gather even more personal data without your knowledge.
Dashlane has explained how attackers managed to download encrypted password vaults by targeting many users at once, increasing their odds of success. The company says no master passwords were compromised, but the incident underscores the importance of strong security practices.
A newly discovered zero-day exploit completely defeats Windows 11's default BitLocker encryption. Microsoft says it's investigating but has not yet provided a fix. The exact mechanism is unclear, but the flaw could leave millions of users vulnerable to data theft.
Google has launched a new feature to detect and warn users about AI-generated deepfake calls. This tool helps protect people from scammers impersonating trusted contacts using AI voice cloning.
Meta's AI support chatbot was used by hackers to take over Instagram accounts. The exploit involved tricking the chatbot into changing account emails and resetting passwords. Meta has since fixed the issue.
The head of GCHQ, the UK's intelligence agency, has warned that AI could become an unstoppable force in cyber warfare. This highlights the urgent need for global cooperation to manage AI's risks and ensure its responsible use.
A new study reveals a simple method to bypass safety filters in AI image generators. The technique, called PAST2HARM, uses past tense prompts to trick models into creating harmful content.
A critical vulnerability called "BadHost" has been discovered in Starlette, a popular open-source package. This affects millions of AI agents and could compromise security.