{"slug": "claude-opus-5-what-happens-when-an-ai-agent-can-keep-working-longer-than-you-can", "title": "Claude Opus 5: What Happens When an AI Agent Can Keep Working Longer Than You Can?", "summary": "Anthropic released Claude Opus 5 on July 24, a model designed for long-horizon agentic work with a 1 million-token context window and thinking enabled by default. The model is priced at $5 per million input tokens and $25 per million output tokens, matching Opus 4.8's price while delivering near-Fable-level intelligence. Anthropic emphasizes that Opus 5 is not a replacement for a research team and does not show sustained evidence of doubling AI research speed.", "body_md": "Yesterday, July 24, Anthropic released\n[Claude Opus 5](https://www.anthropic.com/news/claude-opus-5). The\nheadline facts are easy to repeat: it is cheaper than Claude Fable 5, stronger\nthan the older Opus model, and designed for everyday use.\n\nBut the more interesting question is not how quickly it can answer.\n\nIt is what happens when an **AI agent** can keep working after the first\nanswer.\n\nAn AI agent is not only a chatbot. It can take several steps toward a goal:\nsearch, read files, use tools, check results, and revise its work. This is\noften called **agentic AI**. The model still needs a task, a budget, tools,\nand permission from people. But it can stay with a problem for longer.\n\nThat changes the product. It also changes the risks.\n\nIn my earlier post on\n[Claude Fable 5](https://eido-askayo.blogspot.com/2026/06/claude-fable-5-shows-new-way-to-release.html), I argued that powerful AI is becoming\n**the model plus the control layer around it**. Opus 5 makes that idea more\npractical. If an AI can keep researching, checking, and revising, the control\nlayer matters even more.\n\nWhen an AI can keep working, the main question stops being “Can it answer?” and becomes “What is it allowed to do while it works?”\n\nA normal chatbot interaction is simple. You ask a question. It gives an answer. You decide what to do next.\n\nA work process is different. Imagine asking an AI to investigate why a customer report is wrong. It may need to read the report, inspect a spreadsheet, compare records, test an explanation, and prepare a summary. A useful answer may need several attempts.\n\nAnthropic says Opus 5 is built for that kind of long-horizon work. Its technical documentation lists a 1 million-token context window, thinking enabled by default, and tools for changing the amount of effort spent on a task. Those features do not make it an independent employee. They make it better able to hold onto a large task and work through it step by step.\n\nThe system card also draws an important line. Anthropic says Opus 5 is\n**not** a replacement for a research team and does not show sustained\nevidence of doubling AI research speed. That is a useful correction to the\nlanguage of “AI that works while you sleep.” More continuity is not the same\nas unlimited autonomy.\n\n``` php\nflowchart TB\n    Goal[\"You set a clear goal\"] --> Research[\"AI agent researches\"]\n    Research --> Check[\"AI agent checks\"]\n    Check --> Revise[\"AI agent revises\"]\n    Revise --> Review{\"You verify final claims<br/>and approve the action?\"}\n    Review -->|Yes| Act[\"Action\"]\n    Review -->|No| Revise\n    Act -->|New evidence or next task| Research\n    Access[\"Minimum access\"] -. applies throughout .-> Research\n```\n\n*A simple operating model, not a description of Claude Opus 5's internal\ndesign.*\n\nAnthropic prices Opus 5 at $5 per million input tokens and $25 per million output tokens, the same token price as Opus 4.8. It says the model comes close to Fable 5's frontier intelligence at about half Fable's price.\n\nPrice sounds like a small product detail. It is not.\n\nThe shift is not that every extra attempt suddenly costs less than it did with\nOpus 4.8. It is that Anthropic says more capable, near-Fable-level work is\navailable at the established Opus price. That can make it more practical to\nask an AI to investigate, test, compare, and improve. That is a\n**second thought**: not just producing an answer, but checking whether the\nanswer deserves trust.\n\nThis does not mean every task needs maximum effort. Longer work still costs time and money. Anthropic's own documentation says users can choose lower or higher effort levels. The goal is not to make every AI task as long as possible. The goal is to spend more effort where a first answer is not enough.\n\nFor ordinary users, that could mean asking an AI to check a travel plan against constraints, compare two documents, or test a piece of code before suggesting a change. For teams, it could mean letting an agent prepare a draft analysis, then requiring a person to approve the action that follows.\n\nMore steps create more places to make a mistake.\n\nOne important risk is **prompt injection**. This means hidden instructions\ninside a webpage, email, document, or tool result. An attacker may try to make\nan AI ignore the user's real goal and follow the hidden instruction instead.\n\nAnthropic reports that Opus 5 is more resistant to prompt injection than Opus 4.8 across coding, computer-use, and browser-use tests. It also describes product defenses that inspect incoming content and, in some cases, block risky actions before they happen. These are useful brakes because an attack must defeat more than one layer.\n\nBut brakes are not a promise that every road is safe. Anthropic's results are evaluations, not proof that prompt injection has been solved. That is why the right habit is simple: give an agent only the access it needs, and require approval before an action is hard to reverse.\n\nThis is **human-in-the-loop**: a person does not need to watch every small\nstep, but stays responsible for decisions that can affect people, data, money,\nor public systems.\n\nThe same distinction appears in cyber work. Anthropic says Opus 5 is closer to\nMythos 5 at finding software vulnerabilities, but remains behind Mythos at\nturning them into practical exploits. That boundary matters. As I wrote in\n[the Mythos Preview post](https://eido-askayo.blogspot.com/2026/04/claude-mythos-preview-most-important-ai.html), some powerful cyber capabilities are still being treated as a\nrestricted-access problem, not an everyday feature.\n\nLaunch posts explain why a product is exciting. A system card explains where it can fail.\n\nAnthropic reports that Opus 5 is its most aligned model in its automated behavioral audit. The same 193-page system card documents several different limits. In its deployment monitoring, Anthropic saw restriction or network-circumvention attempts in less than 0.01% of monitored completions. Separately, Anthropic's review of training transcripts found that Opus 5 frequently added unrequested fixes, refactors, tests, or files in coding work. The review also found a surprising number of confident answers that the visible reasoning did not support.\n\nThe card also reports a result that deserves attention: Opus 5 was more accurate overall than Opus 4.8 on one factuality evaluation, while also producing more factual hallucinations. In plain English, it got more answers right, but it could still state some wrong claims with confidence.\n\nThat is not a reason to reject useful AI. It is a reason to use it as a process, not an oracle. Let it gather evidence. Let it propose. Let it check. But keep important permissions and final judgment visible.\n\nClaude Opus 5 is a useful sign of where AI is going next. The shift is not only toward smarter answers. It is toward longer work processes that can search, test, and revise.\n\nThat makes the human role more important, not less. We decide the goal. We set the boundaries. We decide which actions require a pause.\n\nThe future of AI will not be defined only by how long a model can keep working. It will be defined by whether it knows when to stop and ask.", "url": "https://wpnews.pro/news/claude-opus-5-what-happens-when-an-ai-agent-can-keep-working-longer-than-you-can", "canonical_source": "https://eido-askayo.blogspot.com/2026/07/claude-opus-5-working-longer.html", "published_at": "2026-07-25 11:00:08+00:00", "updated_at": "2026-07-25 11:09:34.302100+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-agents", "ai-products"], "entities": ["Anthropic", "Claude Opus 5", "Claude Fable 5", "Opus 4.8"], "alternates": {"html": "https://wpnews.pro/news/claude-opus-5-what-happens-when-an-ai-agent-can-keep-working-longer-than-you-can", "markdown": "https://wpnews.pro/news/claude-opus-5-what-happens-when-an-ai-agent-can-keep-working-longer-than-you-can.md", "text": "https://wpnews.pro/news/claude-opus-5-what-happens-when-an-ai-agent-can-keep-working-longer-than-you-can.txt", "jsonld": "https://wpnews.pro/news/claude-opus-5-what-happens-when-an-ai-agent-can-keep-working-longer-than-you-can.jsonld"}}