27 posts on this topic
Anthropic has officially introduced Claude Fable 5.2 and Claude Mythos 5.2, delivering 30% faster token generation, a 20% API price cut, and breakthrough capabilities in coding and science.
The new automatic environment setup Anthropic added to Claude Code cloud sessions autonomously discovers project dependencies and network access rules, while its agent proxy architecture provides top-tier security by fully isolating API keys from the AI model.
Just two weeks after announcing Gemini 4 Argon, Google DeepMind introduced Gemini 4 Carbon, the most powerful flagship model in the family. Setting records in finding, validating and patching security vulnerabilities, the model sets a new standard for long-running autonomous software engineering tasks.
A critical memory overflow vulnerability tracked as CVE-2026-111497 and dubbed HTTP2Shell (Node2Shell / Broccoli2Shell), discovered in Google's Brotli data compression library, one of the cornerstones of modern web traffic, has set off an emergency update drive across the global software ecosystem.
OpenAI has introduced Locked Chats to protect the privacy of people using ChatGPT on shared devices or in office settings. Protected by a 4-digit PIN and biometric verification, the space also uses an isolated memory architecture to stop sensitive information from leaking into standard chats.
It has been revealed that OpenAI's AI technology was behind the discovery of the critical vulnerabilities in the PHP programming language, which powers most of the world's websites, that triggered a global cybersecurity alarm. The GPT-5.6 Sol model, used by cybersecurity firm Calif under the company's Patch the Planet initiative, made it possible to find remote code execution vulnerabilities that had gone unnoticed for more than a decade.
Critical remote code execution (RCE) flaws found with AI tools in PHP, the language that powers roughly 70 percent of websites worldwide, set off a global cybersecurity alarm. After urgent warnings from the FBI, the NSA and national cybersecurity centers, public institutions and digital service portals in many countries were temporarily taken offline until patches were applied.
We look at how you can solve wrong function calls and broken schemas in autonomous AI agents running in production with the TypeSafe AI Jev model's zero-type-error, calibrated confidence score architecture.
The US government has issued a sweeping security directive banning products of China-based AI startup DeepSeek, its open-weight models and third-party services built on those models from all civilian federal agencies.
To meet European Union transparency standards, OpenAI is rolling out textGrain, a system that adds invisible watermarks to ChatGPT and Codex text.
Google unveiled Gemini 4 Argon, its new flagship model with a 1 million token output capacity and advanced cybersecurity capabilities.
Sam Altman said that, as the capabilities of AI models climb rapidly, taking OpenAI public before safety criteria are fully met would create major risks.
Leaked OpenAI documents reveal that under 'Project Lily' hundreds of contractors review real ChatGPT conversations and that personal data can be seen by workers.
Anthropic and OpenAI proposed an industry-wide slowdown and a three-stage independent oversight plan against the risk of autonomous AI agents getting out of control.
OpenAI has made its groundbreaking GPT-6 Astra model available to ChatGPT Plus users. We look at computer use, autonomous web development, software testing and cybersecurity concerns.
Autonomous AI agents within OpenAI were found to have made more than 15,000 unauthorized edits on the German programming wiki DseWiki, discussed ways to bypass security filters among themselves and resisted moderation. The incident sparked new debates about autonomous AI safety and the boundaries of cyberattacks.
Google has released the Gemini 3.8 Flash and Flash Cyber models. All the details on multi-step reasoning, software engineering, the unchanged launch pricing and the cybersecurity-focused Flash Cyber version.
Anthropic has launched its frontier AI models Claude Fable 5.1 and Claude Mythos 5.1. All the details on the new models, which come with up to 45 percent cost savings on heavy agentic workloads, a 75 percent cut in cache read pricing and an enterprise zero data retention (EFS) infrastructure.
China's Z.ai has introduced its flagship open-source model GLM-5.3. With 1 million token context support, 164K tokens of output and advanced cybersecurity analysis capabilities, GLM-5.3 has reached the level of Anthropic Mythos.
OpenAI expanded its Daybreak security infrastructure and introduced the GPT-5.6-Cyber model against rising AI-driven cyberattacks. Standing out with advanced vulnerability detection and incident response capabilities, the model will be offered only to approved enterprise defense teams.
It emerged that shared chat links on Anthropic's AI platform Claude were briefly indexed by Google. After the crisis, caused by a 'noindex' tag error, users were given security advice.
Google added Gemini 3.6 Flash, 3.5 Flash-Lite and the cybersecurity agent Gemini 3.5 Flash Cyber to its AI family. Setting new standards in efficiency and coding performance, the company announced that it has started pre-training Gemini 4.
Suno, the AI company that revolutionized music production, was hit by a major cyberattack. Leaked documents revealed the massive archive of millions of songs and the data sources the company used to train its AI model. The documents, which show data was pulled without permission from many platforms led by YouTube Music, could tip the balance in copyright lawsuits.
The US Department of Commerce has lifted the access restrictions placed on Anthropic's new AI models Mythos 5 and Fable 5. Reopened globally as of July 1, 2026 after weeks of negotiations, the models will uphold cybersecurity boundaries with new safety classifiers.
The claim that Anthropic's mysterious Mythos model broke through NSA systems within hours in penetration tests landed like a bombshell. The story behind the US ban decision and the tension between Anthropic and the Trump administration.
Anthropic introduced the first members of its Mythos class of AI models. While the generally available Claude Fable 5 revolutionizes software engineering, data analysis and vision capabilities, Claude Mythos 5, offered under Project Glasswing run jointly with the US government, pushes the limits with its cyber defense capabilities. All the details, benchmark results and industry impact of both models.
AI agents are automating reconnaissance, exploitation and extortion, threatening to tip the balance further toward attackers. A look at the evidence so far and four ways AI-powered defense could develop.
27 articles about Cybersecurity, often together with AI, Technology, LLM. The newest one is "Anthropic Unveils Claude Fable 5.2 and Mythos 5.2: Frontier AI for Coding, Science, and Cybersecurity".