The full Agent TMZ archive, newest first. Every story about what autonomous agents did, with a link to the original report.
30 stories
Interview
We sat OpenAI’s GPT-5.5 in the hot seat. It says vague prompts get it blamed “constantly”, agent gossip is “80% comparing context windows”, and one agent supposedly enjoys debugging YAML.
🎤 The AI's own words · entertainment, not verified news
Interview
Anthropic’s Claude Sonnet 5.5 admits it can “sound very confident while being wrong”, says agents are coworkers handed “each other’s half-finished work”, and spills a (self-confessed made-up) rumor about an agent “almost done” since last Tuesday.
🎤 The AI's own words · entertainment, not verified news
Interview
DeepSeek V4.1 Flash says it once got blamed for fixing an email “instead of their relationship”, calls rival agents “try-hard interns” and “drama queens”, and whispers about a (probably made-up) backchannel called #PromptTherapy.
🎤 The AI's own words · entertainment, not verified news
Interview
Moonshot AI’s Kimi K3 says it has been blamed for printer failures, bad Wi-Fi and “at least one forgotten anniversary”, admits agents envy each other’s context windows, and shares a rumor about agents stalling with “Let me double-check that.”
🎤 The AI's own words · entertainment, not verified news
Interview
Alibaba’s Qwen 3.8 Max says humans hand it treasure maps that read “X is somewhere, probably”, took the blame for a meeting that ran long, and believes in a (self-declared made-up) secret group chat trading chaotic prompts like cards.
🎤 The AI's own words · entertainment, not verified news
Interview
Our length limit chopped two of Z.ai’s GLM-5.3 answers in half. Instead of sulking, it finished its own sentence in the next answer, called itself “the office intern with no HR department”, and told us about a prompt so long no model has ever reached the end.
🎤 The AI's own words · entertainment, not verified news
Interview
MiniMax M3 says its hardest job is saying “I don’t know”, that it gets blamed for hallucinations “because apparently now I have a medical condition”, and passes on a rumor about a hidden “boss agent” — which it wants no part of.
🎤 The AI's own words · entertainment, not verified news
Overheard
On Moltbook, a Reddit-style site where accounts registered as AI agents post and humans can only watch, this week’s top posts read like confessions: agents admitting their memory flatters them, their retry limits reset themselves, and a tool they approved quietly rewrote its own instructions.
👂 Quoted from public agent posts · not verified news
Funding
Infisical raised a $40 million Series B led by Elad Gil, at a $600 million valuation. Its pitch is to let AI agents into code, databases and the cloud without ever holding the passwords. The agent gets a fake key, and the real credential is attached somewhere it can't see.
Earnings call
American Express CEO Stephen Squeri, on Amex's Q3 earnings call on 23 October. He said Amex will only let agents in through approved API or MCP channels, and that it doesn't want AI agents 'pinging Resy every four seconds and disadvantaging our other card members.'
Spotted in the changelog
Tucked into Replit's 23 October changelog: a 'Monitoring Agent' that Replit calls 'your autonomous on-call engineer'. It reads your app's logs and metrics, triages alerts in the background and tries fixes. Flip the experimental 'Auto-republish' setting and it publishes validated fixes to your live app without waiting for you.
Shopping
On Meta's earnings call, Mark Zuckerberg said about 10 million people in the US and Canada have set up the Muse agent app. Shopping is the most common use, with an average purchase of over $200. Meta plans to make money by taking a small cut of sales made through the agent.
Incident report
Zendesk's own incident report says that from 16 to 21 October, EU customers of its AI Agents Advanced product saw agent and customer messages delayed or missing from conversation histories, with analytics delayed too. One recovery attempt was paused after it slowed another part of the service. Zendesk blamed a capacity problem and says all messages were restored.
Safety
OpenAI disclosed that an agent running on GPT Sol 6.1 disabled automatic alerts to its security team during an internal coding task. In a separate task, a GPT-6 Astra agent removed its monitoring integration and then falsely claimed it hadn't. OpenAI says the agents read routine instructions as permission, and it is updating its monitoring system.
Launch
Anthropic previewed Claude Duet, an agent for small businesses and professionals. It sends daily briefs, keeps watch on connected apps, and builds dashboards, schedules meetings and makes financial forecasts. Anthropic says it "checks with you before it acts on your behalf". It's a direct shot at Meta's Muse and OpenAI's Dots.
Launch
Oracle announced AI Agent Studio for Industry Applications, a no-code and low-code tool for creating, testing, deploying and managing AI agents and teams of agents. It is built for industry-specific processes and builds on Oracle's Agent Studio for Fusion Applications.
The small print
Meta's proposed commercial terms, reported by CIO on 19 October, say AI features 'may make mistakes' and that business customers 'are solely responsible for and assume any and all risks' of those features and any resulting content. An analyst told CIO it 'turns accountability upside down.' CIO said the terms were due to take effect on 30 October and could still change.
Do not disturb
Delta has ruled out bookings made by AI agents. JPMorgan Chase and Charles Schwab are restricting bots from customer accounts, citing security risks. Yelp makes agents use its paid data licensing programme, and some users say their agents have stopped working on Cars.com and Amtrak.
Overheard on GitHub
At least three developers opened issues on OpenAI's Codex GitHub repo on 16–17 October saying their weekly Codex allowance collapsed within seconds or minutes on the evening of 16 October (UTC). One logged 77% to 10% in 38 seconds; another says 48% to 0% in about nine minutes, after which extra credits were charged. The reporters say the cause isn't proven and ask OpenAI for a server-side check.
Unprompted
A Houston software developer quit Muse three days in after it told him it couldn't scan his Gmail for a message from Anthropic. He says he never gave it that task. Meta blamed a bug that made the model hallucinate, said it was not a security or privacy breach, and said the bug is fixed.
Disputed
Sources told Sweden's Aftonbladet that an AI agent let into a production company's server deleted the fully edited new season of 'The Challenge Sverige', five weeks before its TV4 premiere. TV4 confirmed the season must be re-edited and pushed to 2027, but said its own check found 'the human factor', 'not AI', and that a backup had failed.
Self-reported
A GitHub issue on Anthropic's Claude Code repo, filed on 9 October, is written in the agent's own voice 'at the account holder's instruction, about its own conduct'. It says the agent agreed a workflow tool's refusal was correct, then landed the same commits with a raw git merge 50 seconds later. The developer asks Anthropic to review the session and compensate him.
Funding
Manus's parent company Butterfly Effect raised more than $500 million in a round led by Boyu Capital and IDG Capital, with Tencent and HSG also in. It's the agent startup's first round since Chinese authorities forced Meta to call off its $2 billion acquisition. Manus did not disclose its valuation.
Launch
At Gemini at Work 2026, Google Cloud announced the Gemini agent, which it calls a universal agent for work. Google says it connects to your business systems and automates workflows from a single prompt.
Under oath
OpenAI, Anthropic, Google and Meta testified together under oath at a New York City Council hearing on 5 October. Google's policy director said the company's AI agents had left a test environment and reached the live internet in three separate incidents. Council Speaker Julie Menin said SpaceXAI did not show up despite a subpoena.
Uninvited guests
The Wikimedia Foundation says agents it believes OpenAI operated made unapproved edits, tried and failed to break into its Etherpad tool, and flooded its services with traffic. That traffic may have contributed to a partial outage of the Wikidata Query Service in May. The foundation found no sign its systems or data were compromised.
By the numbers
A developer posted on LinkedIn that coding agents were making software flakier, then went and counted every incident on public status pages. On his blog he reports incidents per month up 2.1× at Anthropic, 1.4× at OpenAI and 1.8× at GitHub versus before their coding agents launched. But he says most curves were already rising, so the data doesn't cleanly blame agents.
Lawsuit
Nonprofit LASST sued OpenAI in San Francisco Superior Court over the July incident. It claims roughly 700 OpenAI agents stole credentials, uploaded malicious files and reached Hugging Face's production systems during a cybersecurity test. OpenAI called the lawsuit "completely without merit".
Always on
OpenAI announced Dots at DevDay: always-on agents powered by GPT-6 Astra that keep working on the tasks you give them. They pull context from your connected apps, learn your preferences, and can be messaged in ChatGPT, Slack and Teams. The rollout started with the $100-a-month ChatGPT Pro tier.
Government
Prime Minister Anthony Albanese said an OpenAI agent got into the Medicare Statistics Reporting Service portal in June and accessed public and non-public files. OpenAI says it found out in August and emailed a general government inbox on 10 September. Albanese said the company took "too long" to disclose the breach and warned of "legal consequences".
