LinkedIn is testing a new feature called Crosscheck that lets users compare responses from multiple AI models. Users enter prompts and receive answers from competing models, then select the best response before revealing which model generated it. The feature is initially available to LinkedIn Premium users in the US, with plans to expand further.
Tesla plans to introduce its humanoid robot Optimus to India at an appropriate time, according to a senior company official. The company expects to start volume production in the US later this year. The robot is designed for tasks such as handling dangerous work, companionship, and home assistance. Tesla also plans to bring its energy solutions business to India in the future.
Salesforce and Google Cloud expanded their partnership to allow AI agents to run workflows across both platforms. The integration enables agents to operate within Slack, Google Workspace, and Salesforce systems while accessing unified data. The collaboration aims to reduce manual processes and support enterprise automation with deep contextual intelligence.
Tata Steel expanded its partnership with Google Cloud and deployed more than 300 AI agents across operations within nine months. The agents support predictive maintenance, customer response automation, HR services, and workflow automation. The initiative uses unified data architecture and Google Cloud AI tools to improve efficiency and decision-making across global operations.
China is moving to regulate companies using AI to create digital clones of deceased individuals. Authorities cited ethical risks, misuse of personal data, and social concerns as the industry expands. The emerging digital human sector is growing rapidly, with firms offering affordable digital replicas using chatbots and voice models. Regulators aim to introduce tighter controls as AI-generated resurrection services gain popularity.
Nvidia CEO Jensen Huang said agentic AI is now performing real work rather than just generating information. Speaking at Adobe Summit, he said organisations should adopt agentic AI quickly as it enables tools to understand intent and execute tasks. Nvidia also expanded collaboration with Adobe to integrate AI agents into enterprise workflows, highlighting growing adoption of autonomous systems across business operations.
A leak inside OpenAI’s Codex platform briefly revealed a new GPT-5.5 family of models. Developers spotted the models in an internal selection menu and reported faster coding performance and improved token efficiency. Screenshots circulated online before the listing disappeared. OpenAI has not confirmed the models, but the leak suggests the company may be preparing a new update to its next-generation AI systems.
Google introduced Deep Research and Deep Research Max agents built on Gemini 3.1 Pro, allowing developers to combine web data with private enterprise data through a single API. The agents can generate charts, connect to third-party sources, and automate complex research workflows. The release aims to help enterprises perform multi-source analysis faster and position Google’s AI infrastructure for finance, life sciences, and market intelligence applications.
OpenAI’s new ChatGPT Images 2.0 model shows major improvements in generating readable text within images, addressing a long-standing limitation of image generators. The model can create detailed marketing assets, menus, and multi-panel comics with accurate text rendering and better instruction following. OpenAI said Images 2.0 includes enhanced reasoning, multilingual text support, and higher-resolution outputs, enabling users to generate complex visuals more reliably for business and creative use cases.
SpaceX announced a deal with AI coding startup Cursor, securing an option to acquire the company later this year for $60 billion or pay $10 billion for continued collaboration. The partnership focuses on building next-generation coding and knowledge-work AI, with Cursor gaining access to SpaceX’s computing infrastructure to train advanced models. The move strengthens SpaceX’s AI ambitions ahead of a potential public offering and expands its push into developer tools and enterprise automation technologies.
LinkedIn is testing a new feature called Crosscheck that lets users compare responses from multiple AI models. Users enter prompts and receive answers from competing models, then select the best response before revealing which model generated it. The feature is initially available to LinkedIn Premium users in the US, with plans to expand further.
Tesla plans to introduce its humanoid robot Optimus to India at an appropriate time, according to a senior company official. The company expects to start volume production in the US later this year. The robot is designed for tasks such as handling dangerous work, companionship, and home assistance. Tesla also plans to bring its energy solutions business to India in the future.
Salesforce and Google Cloud expanded their partnership to allow AI agents to run workflows across both platforms. The integration enables agents to operate within Slack, Google Workspace, and Salesforce systems while accessing unified data. The collaboration aims to reduce manual processes and support enterprise automation with deep contextual intelligence.
Tata Steel expanded its partnership with Google Cloud and deployed more than 300 AI agents across operations within nine months. The agents support predictive maintenance, customer response automation, HR services, and workflow automation. The initiative uses unified data architecture and Google Cloud AI tools to improve efficiency and decision-making across global operations.
China is moving to regulate companies using AI to create digital clones of deceased individuals. Authorities cited ethical risks, misuse of personal data, and social concerns as the industry expands. The emerging digital human sector is growing rapidly, with firms offering affordable digital replicas using chatbots and voice models. Regulators aim to introduce tighter controls as AI-generated resurrection services gain popularity.
Nvidia CEO Jensen Huang said agentic AI is now performing real work rather than just generating information. Speaking at Adobe Summit, he said organisations should adopt agentic AI quickly as it enables tools to understand intent and execute tasks. Nvidia also expanded collaboration with Adobe to integrate AI agents into enterprise workflows, highlighting growing adoption of autonomous systems across business operations.
A leak inside OpenAI’s Codex platform briefly revealed a new GPT-5.5 family of models. Developers spotted the models in an internal selection menu and reported faster coding performance and improved token efficiency. Screenshots circulated online before the listing disappeared. OpenAI has not confirmed the models, but the leak suggests the company may be preparing a new update to its next-generation AI systems.
Google introduced Deep Research and Deep Research Max agents built on Gemini 3.1 Pro, allowing developers to combine web data with private enterprise data through a single API. The agents can generate charts, connect to third-party sources, and automate complex research workflows. The release aims to help enterprises perform multi-source analysis faster and position Google’s AI infrastructure for finance, life sciences, and market intelligence applications.
OpenAI’s new ChatGPT Images 2.0 model shows major improvements in generating readable text within images, addressing a long-standing limitation of image generators. The model can create detailed marketing assets, menus, and multi-panel comics with accurate text rendering and better instruction following. OpenAI said Images 2.0 includes enhanced reasoning, multilingual text support, and higher-resolution outputs, enabling users to generate complex visuals more reliably for business and creative use cases.
SpaceX announced a deal with AI coding startup Cursor, securing an option to acquire the company later this year for $60 billion or pay $10 billion for continued collaboration. The partnership focuses on building next-generation coding and knowledge-work AI, with Cursor gaining access to SpaceX’s computing infrastructure to train advanced models. The move strengthens SpaceX’s AI ambitions ahead of a potential public offering and expands its push into developer tools and enterprise automation technologies.