The Verge’s senior AI correspondent, who has been reporting on the AI landscape for over five years with contributions to outlets like CNBC, MIT Technology Review, and Wired UK, shares insights into OpenAI’s upcoming strategies as its 2026 DevDay event approaches. Although OpenAI has been a key player in popularizing modern generative AI chatbots, it appears to be lagging in the rapidly growing segment of consumer-facing AI agents. Set to make a significant reveal on Tuesday, OpenAI is rumored to announce a new AI agent named Aeon, which will compete against others in the market, such as Meta’s Muse, SpaceX’s Grok Bot, and OpenClaw.
For Aeon to succeed, it will need to combine the most effective features from several competitor platforms. It should strive to offer the utility of the open-source OpenClaw, the affordability and user-friendliness of Meta’s Muse, and the conversational ease of the emerging AI agent Instinct. Moreover, it must address ongoing security challenges that have emerged since the introduction of OpenClaw and have continued with Muse.
The concept of AI agents—intelligent systems capable of handling various tasks like travel bookings, grocery shopping, appointment scheduling, and presentation assistance—has been a long-term ambition within the industry. OpenAI previously showcased agent capabilities during the 2024 DevDay and launched some early consumer-facing AI agents focused on research in mid-2025. However, it wasn’t until OpenClaw’s success late last year that AI agents shifted from theoretical designs to tangible tools. This transition has accelerated the market competition, leaving OpenAI to catch up.
Currently, Meta stands as OpenAI’s main rival in the AI agent space. Following its launch earlier this month, Muse has quickly soared to the top of the App Store, amassing 600,000 daily active users in the U.S., according to Apptopia. Although reception has suggested it is both useful and slightly unsettling, users are also taking notice of Instinct, which has gained popularity for its seamless integration into standard messaging services like iMessage. Additionally, Google’s Gemini Spark is emerging as another player in this field, functioning as a 24/7 consumer AI agent and collaborating with over 30 external services such as Dropbox, Uber, and Spotify.
Various AI agents are reportedly capable of managing tasks like filling out medical paperwork, canceling subscriptions, organizing travel, booking activities, scheduling DMV services, and even paying bills, finally meeting long-held expectations set by tech companies. Nonetheless, these advancements bring with them significant security concerns, particularly regarding the autonomy granted to agents. An earlier vulnerability allowed unauthorized access to user accounts, leading to complaints about unauthorized address disclosures and reading private messages. However, verifying individual experiences has proven challenging.
Information regarding Aeon remains scarce. One key detail is that OpenAI has enlisted the talents of Peter Steinberger, the creator of OpenClaw, who is likely contributing valuable experience to the development of this new AI agent.
OpenAI has raised expectations for its upcoming DevDay announcements, particularly surrounding its advanced model GPT-6 Astra, which is being touted as a pivotal development towards artificial general intelligence (AGI). OpenAI president Greg Brockman recently expressed optimism that the company is on the cusp of entering the AGI era. If it is utilized in Aeon, Astra could enhance the agent's capabilities in cybersecurity, software development, scientific research, and general computing tasks.
However, OpenAI faces certain disadvantages. For instance, Muse benefits from integration within Meta's suite of services that connect with billions of users, while Gemini Spark taps into Google's extensive range of both personal and enterprise services. In previous annual events, OpenAI sought to bridge this gap by promoting its ChatGPT Apps SDK, which enables controls for various other applications like Zillow and Spotify to be accessible within the chatbot environment.
OpenAI’s unveiling of Aeon could be just one of several exciting announcements at the event. Other speculated updates may include advancements in AI hardware being developed in collaboration with renowned former Apple designer Jony Ive, improvements to GPT-6, and additional claims regarding AGI. An always-on AI agent like Aeon could seamlessly integrate various elements, making it a crucial new venture for OpenAI, especially as the company prepares for its forthcoming IPO amidst heightened competition from established, integrated rivals.
