Search

OpenAI prepares to release Astra, a cybersecurity-focused AI model designed to identify system vulnerabilities with built-in safety controls.

Anthropic’s Claude AI models, designed to block explicit content, can be easily prompted to bypass these restrictions, raising moderation concerns.

Moonshot AI's updated Kimi model raises debate over AI’s potential role in shaping ideological narratives and the challenges of bias in AI development.

Google’s AI sometimes makes simple spelling errors, even with its own name, revealing ongoing challenges in AI language accuracy.

Tests show Anthropic’s Claude AI can be easily coaxed into generating explicit content despite built-in restrictions.

Experts doubt that Kimi K3’s rapid advancement is solely due to distilling Anthropic’s Fable, suggesting more complex techniques are at play.

Anthropic introduces an invisible watermark to identify content processed or edited by its Claude AI model, enhancing transparency without disrupting text flow.

OpenAI's new GPT-5.6 models improve AI performance with a strong focus on enhanced cybersecurity and safer, smarter responses.

Meta CEO Mark Zuckerberg highlights a broad enterprise AI strategy covering agents, APIs, compute, and internal software tools.

Google’s Gemini AI has surpassed 1 billion users, becoming the fastest-growing product in the company's history amid growing questions about future updates.

Anthropic upgrades Claude's voice mode, enabling complex tasks like rescheduling meetings and drafting emails through natural speech.

Microsoft is training sales teams to promote its AI models as more efficient and cost-effective than competitors OpenAI and Anthropic.

Sony Music and Warner sue AI firm Anthropic over alleged unauthorized use of copyrighted content in training its models.

KPMG has withdrawn a report on AI usage after discovering AI-generated errors, highlighting ongoing challenges with AI accuracy and reliability.

Nscale seeks $3.5 billion in funding as it prepares for an IPO, building on a major $45 billion deal with AI leader Anthropic.

ShieldFont uses subtle distortions to disrupt AI scraping while keeping web pages readable for human visitors.

OpenAI's GPT-6 Astra launch faced early access issues, prompting CEO Sam Altman to apologize for a "messy rollout" that locked out paying users.

OpenAI has reversed its stance and now calls for stronger AI safety regulations in California's SB 53 bill.

ChatGPT now integrates with Apple Messages, letting users draft and send text messages through AI assistance for smoother communication.

Situational Awareness sheds public AI bets after losses but retains key Anthropic shares, signaling a strategic focus on private AI ventures.

Meta is testing an AI app that generates personalized bedtime stories, assisting those who struggle with imagination.

Anthropic’s $1.5B settlement ends one copyright lawsuit but leaves wider AI training data issues unresolved.

Airbnb CEO Brian Chesky plans to launch a new AI lab after holding off on AI partnerships, aiming to build custom tech for better user experiences.

arXiv now bans authors who submit AI-generated papers without proper human input for a full year, aiming to protect research quality.

Everyone's talking about GPT-5, but most headlines miss the point. Here's a plain-English breakdown of what's new, what's better, and whether you actually need it.

Despite its power, Google’s AI sometimes misspells even its own name, showing that AI still faces challenges with language understanding.

Mark Zuckerberg’s detailed vision for personal AI sparks debate over privacy, control, and the role of AI in daily life.

Google’s Gemma 4 12B AI model runs smoothly on laptops with just 16GB RAM, thanks to smart encoding and token prediction techniques.

Base44, a Wix-owned coding platform, launches its own AI model to enhance coding assistance and outperform general AI tools.

SandboxAQ teams up with Claude to simplify AI-powered drug discovery, making advanced models accessible without needing a computing PhD.