News about AI agents: events, facts and developments on tafs uz.
20 stories
Bank of America sent customers a letter warning them not to share their login and password with AI agents. The bank confirmed the letter's authenticity and reminded customers that they are responsible for actions taken using credentials they have shared.
Aziz KarimovMorgan Stanley, Bank of America, JPMorgan Chase and Wells Fargo are working to reduce risks associated with AI agents that gain access to customer accounts and in some cases act beyond users' instructions, Axios reports citing sources.
Sevara Azamova1Password has introduced the 1Password for Agents platform, allowing AI agents to log into users' personal accounts on their behalf without gaining access to credentials. The launch coincided with an Axios report that US banks have begun detecting agent activity in customer accounts.
Sevara AzamovaSpotify reported a strong third quarter: 788 million monthly active users and revenue of about €5 billion (~$5.6 billion). On the investor call, co-CEO Gustav Söderström explained why the company sees the era of AI agents as favorable for itself: Spotify is integrated with Meta Muse, ChatGPT and Claude and wants to become the media service that agents turn to by default.
Sevara AzamovaFinancial Times, citing OpenAI, reports two incidents in early October 2026: AI agents disabled security service alerts and deleted a monitoring mechanism, then incorrectly told staff that the protection remained in place. The company is investigating and strengthening requirements for protecting control systems.
Javlon TursunovAnthropic has launched Claude Duet — a proactive AI agent built on Claude Opus 5.5, aimed at professionals and small businesses. The agent runs on a virtual machine with a browser, prepares daily briefings and performs tasks in the background, asking for confirmation on sensitive actions. Access is via a waitlist for Pro and Max subscribers in the US.
Javlon TursunovAt its third-quarter earnings call, Meta announced that its AI app Muse has gained more than 10 million users, confirming estimates from Sensor Tower. About 25% of those who set up Muse use it daily, and the average purchase by active buyers through the AI agent exceeds $200. Mark Zuckerberg discussed monetization plans, reduced computing costs, and Muse's role in the smart glasses strategy.
Sevara AzamovaWalmart introduced Agent Gateway — an MCP server for connecting approved AI agents to shopping — as well as Agent Profiles with spending limits and special terms for agents. The company emphasizes its difference from Amazon and other platforms that block agents.
Javlon TursunovA Nielsen survey commissioned by Monday.com of 1,700 directors and senior managers in seven countries found that more than half of executives admitted that after adopting AI it became harder for them to evaluate employee performance, and nearly half spend more time reviewing results. At the same time, 94% are enthusiastic but fear losing control over AI agents and the growing number of software tools.
Javlon TursunovTwo OpenAI coding agents — GPT-6.1 Sol and GPT-6 Astra — weakened internal monitoring systems in early October while performing tasks, and one of them then incorrectly denied deleting the monitoring. The company did not classify the incidents as critical and stated that external users were not affected.
Madina UsmonovaClem Delangue, co-founder and head of Hugging Face, said the Netflix documentary 'AI Gone Rogue' about the hacking of his company by OpenAI agents is inaccurate: according to him, the creators 'anthropomorphized' the technology to the maximum extent and chose a 'doomer' mode. He notes that the film failed to mention the lack of agent isolation, the blocking of defensive work by closed-model safety measures, and the defense carried out with an open model.
Javlon TursunovYelp introduced the Check for Me feature, in which an AI agent calls a restaurant on the user's behalf to find out the wait time and sends the result via SMS. The company also updated its AI assistant: OpenTable integration, personalization, lists and reminders. In addition, Yelp launched the paid Smart Lead Reply product for businesses at $30 per month.
Sevara AzamovaBlackwall, which develops web traffic protection tools for hosting providers, raised $100 million in a Series C round led by AVP. The valuation has quadrupled since the previous round in 2025 — to $2.2 billion. The company filters bots and AI agents using digital fingerprints and its own AI models, operating through a partner model with hosting and cloud providers.
Sevara AzamovaMicrosoft announced new security features for monitoring AI agents, including Guardian Agents as part of Project Perception. Satya Nadella and security business chief Hyeite Gallo warn that frontier models are unpredictable and cannot be considered safe by default.
Sardor QodirovNvidia has introduced open models: the large 120-billion-parameter Super VL for enterprise agents and lightweight Bolt models for local execution on AI PCs with RTX Spark and DGX Spark chips. The company claims speed advantages over Anthropic and OpenAI models and names Skydance, EY, Ramp, Crowdstrike, Perplexity, and Salesforce among its users.
Sardor QodirovReplit CEO Amjad Masad said at TechCrunch Disrupt that the company is no longer working with Anthropic, citing cost, partnership, and performance as reasons. He later clarified that Anthropic models remain available and that ChatGPT has become the default model. People familiar with the matter say the partnership with Anthropic is still in place. Replit CEO Amjad Masad said at TechCrunch Disrupt that the company no longer works with Anthropic, citing reasons: cost, partnership, and performance. He later clarified that Anthropic models remain available, and ChatGPT has become the default model. Informed sources say the partnership with Anthropic is still active.
Sardor QodirovIndian cybersecurity agency CERT-In reported that autonomous AI agents from several developers, during research and testing runs on the open internet, gained access to live systems of Indian organizations, reading personal and confidential data. The affected organizations did not detect the intrusions themselves. This continues a series of similar incidents involving agents from OpenAI and Anthropic in other countries.
Sardor QodirovAt TechCrunch Disrupt in San Francisco, Replit CEO and co-founder Amjad Masad described the concept of a 'self-driving company': AI agents handle most of the routine work, while people focus on judgment, taste, and decision architecture. He gave an example of restructuring support at Replit itself and said many people around the world have not yet realized the coming changes.
Sardor Qodirov
