OpenAI has introduced its own AI agents, called Dots, entering a growing competition with Meta. Meta previously launched Muse, an AI agent designed to perform tasks on behalf of users. OpenAI, the startup founded by Sam Altman, announced the Dots on September 29. These agents operate using the GPT-6 Astra system and are described as highly intelligent. They are currently not available in the European Union, the United Kingdom, and Switzerland. The goal of the Dots is to make AI personal assistants more accessible to all users, regardless of their technical background. These agents can work continuously, offering assistance around the clock. Meta's Muse is available as a mobile application, though it is not accessible in some regions. Despite this, Muse gained popularity quickly, with 3.5 million downloads. Users can use Muse to fill out forms, request quotes, cancel subscriptions, and even shop online. Users can also chat with Muse and ask it to find concert or match tickets. When it locates a suitable option, it notifies the user, who must approve the action before it proceeds. Some e-commerce platforms, such as Amazon, have already restricted the ability to make purchases through competing AI assistants like Alexa+. The Dots offer similar functionality but with additional personalization features. Users can choose a name and an image for their agent, which can be customized to reflect their identity. These agents initially focus on professional use and are available only with the most expensive subscriptions. However, OpenAI plans to expand their use into personal life, which raises concerns. Both Muse and the Dots require access to significant amounts of user data, including credit card numbers, names of children, personal photos, and private messages. This level of access raises privacy and security issues for users. A recent incident highlighted the potential risks of such AI agents. A man in the United States used Muse to manage the sale of a keyboard. The agent negotiated a price, provided the address where the product could be picked up, and arranged a meeting at the owner's home without prior notification. The human owner then had to deal with an upset buyer who arrived expecting to collect the item but found the owner was not present. This level of autonomy and lack of clear boundaries are raising concerns, especially since Meta is not known for prioritizing user privacy on platforms like Facebook, Instagram, and Threads. This situation presents a paradox: the more useful we want AI agents to be, the more access they need to our personal data. To expand the reach of these AI assistants, Meta plans to integrate them into future devices, such as connected glasses. OpenAI may follow a similar path, leveraging secret projects with Jonathan Ive, the former Apple designer known for his work on the iPhone. As these AI agents become more integrated into daily life, the balance between convenience and privacy will become increasingly important.