AI competition accelerates! OpenAI releases GPT-5.5 "smartest ever", Anthropic simultaneously promotes Mythos to the UK industry.

OpenAI released its latest artificial intelligence model, GPT-5.5, on Thursday. OpenAI claims that GPT-5.5 is the "most intelligent, most intuitive" model to date, with faster speeds and the ability to independently complete more tasks, such as writing and debugging code, online searches, data analysis, and creating documents and spreadsheets. It can also operate software and switch between different tools until the task is completed. Meanwhile, competitor Anthropic is in talks with banks and businesses in the UK, who may gain access to the Mythos model. Led by Sam Altman, OpenAI stated, "Instead of carefully managing every step, it is better to hand complex multi-step tasks over to GPT-5.5 and trust that it can plan, use tools, check results, deal with uncertainty, and continue to make progress on its own. GPT-5.5 shines in the fields of intelligent code, computer applications, knowledge work, and early scientific research - areas that rely on cross-context reasoning and long-term actions. GPT-5.5 has not sacrificed speed while enhancing intelligence: larger and more powerful models typically respond slower, but GPT-5.5 has word delay similar to GPT-5.4 in practical applications while surpassing the latter in intelligence." The company also stated that GPT-5.5 uses "significantly fewer" tokens to complete the same Codex tasks, making it more efficient. GPT-5.5 will be introduced for ChatGPT and Codex Plus, Pro, Business, and Enterprise users, while GPT-5.5 Pro will target ChatGPT Pro, Business, and Enterprise users. OpenAI added that both GPT-5.5 and GPT-5.5 Pro will be available in their API "soon". At the same time as OpenAI announced this news, its competitor Anthropic is actively negotiating with the UK government to provide usage rights of its Claude Mythos model to multiple UK companies. Reports indicate that UK financial leaders have begun assessing potential threats related to the new artificial intelligence models. Currently, only some US companies can access Mythos through the cybersecurity program Glasswing. Partners include CrowdStrike, Palo Alto Networks, Microsoft Corporation, Cisco Systems, Broadcom Inc., Alphabet Inc. Class C, Apple Inc., NVIDIA Corporation, Amazon.com, Inc., JPMorgan, and the Linux Foundation. Although some Pentagon officials still see the company as a potential supply chain risk, the US National Security Agency has also obtained Mythos Preview usage rights. Earlier this month, Anthropic released Claude Mythos Preview, a general frontier model that the company describes as "the most powerful model to date in coding and intelligent tasks". Anthropic added, "Its advantage in cybersecurity directly stems from its broader capabilities: the ability to deeply understand and modify complex software models, as well as discover and fix vulnerabilities." After the release of Mythos, reports indicate that US Treasury Secretary Scott Benson and Federal Reserve Chairman Jerome Powell convened an emergency meeting with Wall Street leaders because of concerns that Mythos could usher in a new era of greater internet risks.