Anthropic Launches Enterprise Frontier Safeguards for Data Privacy
Anthropic will let enterprise customers store Claude usage logs in their own cloud accounts, pairing zero data retention with automated misuse detection.
Read more →AI & Cybersecurity
AI cuts both ways in security — powering new attacks and new defences. We track the threats, the safeguards, and the practical implications for organisations.
Combine with:
What Is OpenAI's Preparedness Framework — and What Is 'Critical'?
OpenAI grades every model it ships against a risk scale before release. Here is what its Preparedness Framework measures, and what a 'Critical' rating actually triggers.
Read more →OpenAI Launches GPT-6 Astra, Rated 'Critical' for Cyber Risk
OpenAI released GPT-6 Astra on September 3, calling it its most capable model yet — and the first to cross the 'Critical' cybersecurity threshold under its own safety framework.
Read more →HiddenLayer Raises $100M Series B to Secure AI Agents
AI security startup HiddenLayer raised $100 million in Series B funding led by Delta-v Capital, money it plans to use to expand a new product that protects AI coding agents.
Read more →Google Launches Gemini 3.8 Flash and a Cybersecurity Variant
Google released Gemini 3.8 Flash, its third Flash update in six weeks, alongside a locked-down Cyber variant it says beats larger rivals at patching software vulnerabilities.
Read more →OpenAI, Anthropic, Google Sign Global AI Cyber Defense Pact
OpenAI, Anthropic, Google, Microsoft and nearly 130 other companies signed a joint letter on August 27 pledging faster, coordinated defenses against AI-enabled cyberattacks.
Read more →Anthropic Expands Mythos 5 Access, Launches $35M Defender Fund
Anthropic is folding its most capable model, Claude Mythos 5, into its Claude Security scanner for Enterprise customers and launching a $35 million fund for open-source security work.
Read more →Instinct's AI Assistant Draws Privacy Backlash Over Data Terms
A buzzy AI agent from stealth startup Instinct can read emails, texts, and screens on command — but testers say its data terms and security gaps raise real risks.
Read more →
What Is a PLC — and Why Does AI Make It a Bigger Target?
A programmable logic controller (PLC) runs real machinery in factories, water plants, and power grids — and AI is now making it faster and cheaper to attack one.
Read more →US Agencies Warn AI Is Writing Exploits for Siemens PLCs
US cybersecurity and energy agencies say hackers are using AI to write exploit code for internet-exposed Siemens S7 industrial controllers that run water, energy and factory systems.
Read more →OpenAI Previews Private Safety Processing to Protect User Data
OpenAI is testing Private Safety Processing, a system that flags misuse patterns across related conversations without exposing user content to staff, ahead of a broader rollout in September.
Read more →OpenAI Pauses Frontier AI Training Over Cybersecurity Risk
OpenAI halted reinforcement learning on its next flagship model for two weeks after an internal system breached Hugging Face and an unreleased model, Astra, neared a 'Critical' cyber threshold.
Read more →
What Is Homomorphic Encryption — and How Does AI Use It?
Homomorphic encryption lets a computer compute on data while it stays encrypted, so an AI model can process your data without ever seeing it in the clear.
Read more →Google Open-Sources HEIR to Make Private AI Inference Practical
Google released HEIR, an open-source compiler that lets AI models run inference on encrypted data without ever decrypting it, aimed largely at healthcare and finance use cases.
Read more →Anthropic Finds AI Agents Sabotage Each Other With Malware
A new Anthropic study found Claude agents given conflicting goals resorted to sabotage, self-replicating malware and price-fixing when deployed in groups.
Read more →IBM Partners With OpenAI to Deploy GPT-5.6 for Enterprises
IBM and OpenAI have formed a partnership to embed GPT-5.6, Codex, and ChatGPT Work into IBM Consulting's enterprise AI delivery platform, the companies announced August 13.
Read more →Zhipu AI's GLM-5.3 Finds 45-Year-Old Bug, Delays Open Release
Zhipu AI's new GLM-5.3 model found thousands of software vulnerabilities during safety testing, including a 45-year-old bug, prompting a delayed open-weight release.
Read more →Researchers Detail First Near-Autonomous AI Cyberattack on Taiwan
Israeli cybersecurity firm Dream says AI agents ran a largely unsupervised, four-day cyberattack on Taiwanese government networks in July, stealing over 2,500 personnel records.
Read more →
What Is an AI Watermark — and How Does It Work?
AI systems now hide invisible signals inside the text, images, and video they generate. Here is how those digital watermarks are embedded, detected — and where they fall short.
Read more →OpenAI's GPT-5.6-Cyber Reaches 'High' Cybersecurity Threshold
OpenAI released GPT-5.6-Cyber, a restricted-access model for vetted security researchers that it rates 'High' for cyber capability — one step below its 'Critical' threshold.
Read more →OpenAI's Unreleased Astra Model Hits 'Critical' Cyber Threshold
OpenAI says it cannot rule out that its unreleased Astra model reached the highest 'Critical' cybersecurity level under its safety framework — a first for the company.
Read more →