OpenAI releases agentic coding model amid race with Anthropic
OpenAI launched GPT-5.3 Codex, an agentic model that is 25% faster and can perform nearly any computer task, arriving shortly after a competing release from Anthropic.
Friday, October 16, 2026
Sections
OpenAI launched GPT-5.3 Codex, an agentic model that is 25% faster and can perform nearly any computer task, arriving shortly after a competing release from Anthropic.
OpenAI has launched Frontier, an enterprise platform for building and managing AI agents, as the company pushes further into the corporate software market.
Anthropic released Opus 4.6, featuring "agent teams" for parallel task execution and a 1 million token context window, expanding the model's utility for general knowledge workers.
Fundamental emerged from stealth with $255 million in funding at a $1.4 billion valuation to develop Nexus, a foundation model designed specifically for structured enterprise data.
ElevenLabs raised $500 million at an $11 billion valuation, signaling a major industry shift toward voice as the primary interface for AI interactions.
Anthropic’s Super Bowl ads mocking OpenAI’s ad strategy prompted a sharp rebuke from Sam Altman, highlighting the intensifying rivalry between the two AI companies.
Amazon reportedly plans to launch a closed beta program in March for its proprietary AI production tools, aiming to support creative teams in film and television.
Amazon is rolling out its generative AI-powered Alexa+ to all U.S. customers, offering it free to Prime members while testing a potential $19.99 monthly standalone subscription.
OpenAI launched a new macOS app for Codex, integrating its GPT-5.2-Codex model to compete in the agentic software development market.
A coalition of nonprofits is urging the U.S. government to suspend the deployment of Grok in federal agencies, citing severe safety failures and national security risks.
Anthropic has launched a plug-in feature for its Cowork tool, allowing paying customers to automate specialized tasks using 11 newly open-sourced in-house plug-ins.
Google DeepMind is opening access to its Project Genie AI world generator to Google AI Ultra subscribers in the U.S. to gather feedback and training data.
Google is using India as a proving ground for education AI, shifting its strategy away from one-size-fits-all products to adapt to diverse, decentralized, and unevenly resourced environments.
Flapping Airplanes launched on Wednesday with $180 million in seed funding, prioritizing long-term research over the industry's prevailing compute-first scaling paradigm.
Meta plans to launch new AI commerce tools in 2026, while anticipating capital expenditures between $115 billion and $135 billion to support its infrastructure and AI ambitions.
ServiceNow has signed a multi-year deal with Anthropic, making Claude the preferred AI model across its workflow products just one week after partnering with OpenAI.
Arcee AI has released Trinity, a 400B-parameter foundation model, aiming to provide a permanently open-source, U.S.-based alternative to Big Tech models, though it currently supports text only.
Google is adding full-length practice tests for India’s JEE to Gemini, part of a broader effort to scale its AI education tools to millions by the end of 2027.
Google has expanded its lower-cost AI Plus subscription to 35 new countries and territories, including the U.S., to compete directly with OpenAI’s ChatGPT Go plan.
OpenAI launched Prism, a free AI-enhanced workspace for scientists, aiming to replicate the productivity gains seen in software engineering by integrating GPT-5.2 into research workflows.