AI news.
Understand the developments.

Coforge joins SpaceXSI partner network for enterprise Grok adoption

Coforge announced that it has joined the SpaceX Super Intelligence Partner Network as a Transformation Partner. It plans to help enterprises integrate the Grok portfolio into existing systems and business workflows through its Momentuum blue engineering unit. This is a partnership announcement, not evidence of completed customer deployments or independently measured savings.

Turn this topic into a learning task →

Google announces Gemini Omni 1.2 Flash for longer video clips

Google announced Gemini Omni 1.2 Flash with native 15-second video generation, conversational editing and improved reference following. The company says it is generally available through the Gemini API and Gemini for business, with lower generation pricing than Omni 1.1. Consistency and performance improvements are Google’s claims; account access and the quality of a particular result still need checking.

Turn this topic into a learning task →

OpenAI introduces GPT-6 Cyber for authorised security research

OpenAI introduced GPT-6 Cyber, a specialised model for finding, validating and helping fix software vulnerabilities within an authorised scope. Its system card describes restricted access through Daybreak Red and rates its cybersecurity capability as Critical under OpenAI’s own Preparedness Framework. Reported benchmark gains are the company’s evaluation results, not a guarantee of safety or effectiveness in every deployment.

Turn this topic into a learning task →

Official system card and access safeguards ↗

NVIDIA adds four Cosmos 3 checkpoints for interactive simulation

NVIDIA researchers published four new open-weight Cosmos 3 Nano checkpoints for simulated driving scenes, camera and LiDAR observations, simulator-to-video generation and robot-action video prediction. The release is intended for developers working on world models and physical AI. Its authors explicitly say that the fidelity and limits of these simulations still need measurement.

Turn this topic into a learning task →

LTM and ServiceNow plan wider enterprise agentic AI partnership

LTM and ServiceNow announced a planned expansion of their partnership to help customers adopt governed agentic AI across business functions. The release says LTM already uses 1,500 ServiceNow AI agents internally and will use that experience as a reference for joint solutions. Deployment counts and efficiency expectations are company-reported; the partnership expansion remains a plan.

Turn this topic into a learning task →

Workplace AI monitoring raises privacy concerns

Australian researchers call for stronger safeguards as employers increasingly monitor staff.

Turn this topic into a learning task →

Why Indian enterprises struggle to scale AI

In an interview, IBM’s Sriram Raghavan points to data quality, security and skills as barriers to moving from experiments to wider deployment.

Turn this topic into a learning task →

US civilian agencies directed to remove DeepSeek technology

The newspaper reports a Department of Homeland Security order to remove DeepSeek technology from federal civilian IT systems and restrict future purchases.

Turn this topic into a learning task →

MiniMax announces M3.1 Flash

MiniMax says its updated model improves agentic tasks, programming and cybersecurity while reducing usage costs. These are company claims, rather than independent benchmark findings.

Turn this topic into a learning task →

The AI investment boom complicates inflation policy

Reuters examines how heavy AI investment can add price pressures while higher borrowing costs affect other businesses and households differently.

Turn this topic into a learning task →

China proposes controls on sensitive AI technology exports

Proposals cover certain models, agents and computing technologies. Reported exceptions distinguish overseas technology transfers from access through APIs and online services. The measures are proposals, with consultation continuing.

Turn this topic into a learning task →

India’s trust and safety debate puts AI and children in focus

A New Delhi panel discussed deepfakes, child protection, privacy and whether voluntary testing is sufficient. Participants also considered safety evaluations suited to India’s diverse cultural context.

Turn this topic into a learning task →

AI self-replication warning draws concern and scepticism

Anthropic described experiments and potential risks involving model replication. Experts differed on the implications; controlled tests do not establish that the projected internet-wide scenario will occur.

Turn this topic into a learning task →

India plans an AI regulation consultation paper

IT Minister Ashwini Vaishnaw said the government would release a consultation paper, focusing on safety, deepfakes, skilling and inclusion. An announced consultation is distinct from enacted regulation.

Turn this topic into a learning task →