Moonshot AI says its 2.8 trillion-parameter Kimi K3 matches or outperforms leading proprietary models in several coding and software engineering benchmarks, highlighting the rise of open-weight AI
Check Point Research warns that AI-generated faces, voices and identity documents are making remote identity verification less reliable, forcing banks to rethink digital trust and KYC security
Mozilla says open-weight AI has nearly matched closed models in capability and costs less, but businesses still favour proprietary systems because deployment, support and governance remain harder
PrismML says Bonsai 27B compresses Alibaba's Qwen3.6 27B model enough to run locally on phones and laptops, highlighting how low-bit quantisation could expand on-device AI
Proprietary and open-weight AI represent two competing approaches to building and commercialising artificial intelligence. Here's how they differ in technology, deployment, costs and business strategy
AI hallucinations are no longer merely producing wrong answers. Attackers are registering fake domains and software packages that models may recommend, creating new routes for phishing and malware
AI assistants are beginning to remember users' conversations, preferences and digital context. Here's how AI memory works, why companies are adopting it, and what it means for privacy
More than half of Indian users now rely on AI assistants and chatbots for information, with many using them alongside traditional search methods, a YouGov survey has found
Google DeepMind chief Demis Hassabis says AGI could arrive within years and proposes a US-led Frontier AI Standards Body as the basis for an international framework to govern advanced AI models
As frontier AI advances faster than existing safeguards, technology leaders and experts are calling for stronger rules on testing, accountability, cybersecurity and autonomous systems
Apple claims OpenAI targeted engineers working on Apple's hardware programmes and encouraged them to share confidential information after they joined the AI company
Meta's Muse AI let users generate images using public Instagram profiles. Days after launch, the company withdrew the feature following criticism over consent, privacy and impersonation risks
AI is becoming central to the smartphone experience, with 71 per cent of Indian users reportedly using generative AI features and 59 per cent considering AI when buying a new device
As companies build AI-powered in-house systems, India's IT services firms risk losing lucrative software maintenance contracts that have long underpinned the country's outsourcing industry
IBM forecast second-quarter revenue below estimates, saying clients shifted spending towards AI infrastructure, triggering a broader selloff in software stocks
Nasscom's AI-native index finds nine in 10 engineering graduates are AI-native or AI-proficient, but identifies gaps in orchestration and technical grounding
Anthropic is reportedly localising Claude subscriptions for Indian users with rupee pricing, even as UPI payments remain unavailable and expansion efforts gather pace
India is pursuing access to Anthropic's Mythos AI while using alternative models in a CERT-In sandbox to detect software vulnerabilities and strengthen cybersecurity
As enterprises adopt AI, the challenge is shifting from protecting data to safeguarding institutional knowledge. Here's why governance, portability and AI ownership are becoming critical
