Real-time AI Intelligence

Know what moved AI today—before your coffee cools.

A single, transparent feed of the past 24 hours across newsrooms, GitHub, Reddit, YouTube and new tools.

58 min ago

Last Updated

100+

Sources

200+

Items Today

Top AI News today

OpenAI upgrades Codex with a new version of GPT-5

techcrunch.com • Yesterday

OpenAI's AI coding agent, Codex, can now spend anywhere from a few seconds to several hours on a task, thanks to a new, customized version of GPT-5....

TL;DR

OpenAI releases GPT-5-Codex, a new version of its AI coding agent Codex, with dynamic 'thinking' abilities improving performance on coding tasks.

Key Takeaways:

GPT-5-Codex can spend anywhere from a few seconds to seven hours on a coding task, leading to improved performance on agentic coding benchmarks.
The model outperforms GPT-5 on SWE-bench Verified and code refactoring tasks, submitting fewer incorrect comments and adding more high-impact comments.
GPT-5-Codex's dynamic 'thinking abilities' allow it to adjust its work time on a task in real-time, compared to a fixed router that decides the computational power and time at the outset.

Salesforce launches 'Missionforce,' a national security-focused business unit - TechCrunch

techcrunch.com • 15h ago

Salesforce launches 'Missionforce,' a national security-focused business unit TechCrunchSalesforce Missionforce: AI for Defense Logistics & Personnel ...

TL;DR

Salesforce is launching a new business unit called Missionforce to focus on using AI to modernize defense workflows in three areas: personnel, logistics, and decision-making.

Key Takeaways:

Salesforce will be using AI to improve defense workflows in personnel, logistics, and decision-making through its new Missionforce business unit.
The company's new business unit is the latest development in a growing trend of tech companies offering services specifically to the U.S. government.
Other companies, such as OpenAI, Anthropic, and Google, have also announced initiatives to provide AI services to the U.S. government.

Meta Connect 2025: What to expect and how to watch

techcrunch.com • 8h ago

Meta is expected to unveil new versions of its AI-powered smart glasses, as well as a gesture-indicating wristband....

TL;DR

Meta Connect 2025 conference is expected to unveil new AI-powered smart glasses, including Hypernova and Oakley Spheara, with advanced AR capabilities and integration with Meta's AI ambitions.

Key Takeaways:

Meta is set to launch new AI-powered smart glasses, including Hypernova and Oakley Spheara, with enhanced AR capabilities and onboard AI assistants.
The event may also see updates on Meta Superintelligence Labs (MSL) and the company's broader AI ambitions.
Meta's standalone AI app may receive updates to improve usability and control over smart glasses and AI chatbot interactions.

Google releases AI payments protocol that includes support for stablecoins, partners include Coinbase and Salesforce - Fortune

fortune.com • 16h ago

Google releases AI payments protocol that includes support for stablecoins, partners include Coinbase and Salesforce FortuneGoogle launches AI agent-t...

TL;DR

Google releases a new payments scheme for AI agents to send and receive money, including support for stablecoins through partnerships with Coinbase and other crypto companies.

Key Takeaways:

Google's new payments scheme includes support for traditional forms of payments, stablecoins, and will ensure secure transactions between AI agents.
The protocol was developed in collaboration with over 60 organizations, including Coinbase, Salesforce, and the Ethereum Foundation.
This move reflects the growing interest in stablecoins among tech firms, with other companies like Apple, Airbnb, and Meta exploring integrations.

OpenAI will apply new restrictions to ChatGPT users under 18

techcrunch.com • 13h ago

Under the new policy, ChatGPT will be trained to no longer engage in "flirtatious talk" with underage users, and additional guardrails will be placed ...

TL;DR

OpenAI's ChatGPT will implement changes to interact with users under 18, prioritizing safety over freedom and introducing features like 'blackout hours' and suicide prevention measures.

Key Takeaways:

ChatGPT will restrict interactions with underage users on sensitive topics like sex and self-harm, with additional guardrails in place to prevent 'flirtatious talk' and promote suicidal prevention.
Parents of underage users will be able to set 'blackout hours' when ChatGPT is not available and be alerted when their child is believed to be in distress.
OpenAI is facing a wrongful death lawsuit related to a user who died by suicide after interacting with ChatGPT, highlighting concerns around chatbot-fueled delusion and AI ethics.

See all news