Fable 5 asks businesses to think bigger about AI agents

Anthropic has reportedly released Fable 5, a new frontier model designed for long-running, autonomous tasks. The key change isn't just improved capability, but a fundamental shift in how we're expected to use AI. Instead of discrete prompts for small jobs, Fable 5 is built for delegation, asking businesses to imagine and assign complex workflows that an AI agent can manage over hours or even days.

According to the AI Daily Brief, Anthropic's release of a new model, supposedly named Fable 5, marks a turning point in the development of artificial intelligence. While details remain emergent, the core message is a departure from the current paradigm of using AI for discrete, contained tasks.

Fable 5 is reportedly designed for agentic behaviour—the ability to pursue complex goals autonomously over extended periods, potentially for hours or days. This moves the user's role from a 'prompter' who asks for specific outputs to a 'delegator' who assigns high-level objectives.

This shift can be summarised as a move from task-based AI to agentic AI.

Accompanying the launch are reports of some community backlash against the model's 'guardrails', suggesting Anthropic is continuing its focus on safety, potentially at the cost of some flexibility. There are also mentions of emerging 'enterprise retention concerns' and hints that rival OpenAI may be preparing a competitive response, indicating this agentic shift is a new front in the AI race.

For business owners and operators, the emergence of models like Fable 5 isn't just an incremental upgrade. It signals a fundamental change in how we should think about integrating AI into our organisations. The focus is no longer on simply making individual tasks faster, but on automating and redesigning entire workflows.

A shift from tool to team member

The prevailing use of AI in business has been as a sophisticated tool—a better search engine, a faster copywriter, a tireless analyst. Agentic AI, however, functions more like a junior team member you can delegate a complex project to. The mental model shifts from "what can I ask it?" to "what can I assign it?" This requires a higher level of trust and a different style of management, focused on setting clear goals, providing context, and defining constraints, rather than step-by-step instructions.

Workflow redesign becomes critical

You cannot simply 'drop' an AI agent into a legacy business process and expect transformative results. To leverage this technology, businesses must proactively re-engineer their workflows.

• Before: A sales manager manually researches 10 leads, drafts 10 personalised emails, sets follow-up reminders, and logs activities in a CRM. This involves multiple applications and significant context-switching.
• After: The sales manager gives an AI agent a high-level goal: "Identify 10 potential leads in the manufacturing sector in Victoria, draft and send initial outreach emails based on their company's recent announcements, and schedule follow-ups for any who respond positively. Log all activity in our CRM and provide me with a summary report at the end of the day."

This level of delegation frees up the human operator for higher-value strategic work, such as closing deals or managing client relationships.

Cost, productivity and talent implications

The move towards agentic AI will have direct operational impacts:

• Productivity: The potential for productivity gains is an order of magnitude greater than with task-based AI. Automating multi-day processes could fundamentally change the economics of knowledge work.
• Cost: While the subscription or API costs for such powerful models will likely be substantial, the return on investment could be immense if they can reliably execute workflows that currently take dozens of human hours.
• Talent: The 'enterprise retention concerns' noted in the brief are significant. While agentic AI may automate certain roles, it also creates a demand for new skills. The most valuable employees will be those who can effectively design, delegate to, and supervise AI agents. This 'AI orchestration' will become a critical business capability.

This technology is moving from the lab into commercial reality. While Fable 5 may not be available to everyone immediately, the trend is clear. Proactive businesses should begin preparing now.

1. Audit for agentic potential. Go beyond identifying small, repetitive tasks. Map out your core business processes, such as lead generation, client onboarding, or financial reporting. Ask: "Which of these multi-step, objective-driven workflows could be delegated if I had a tireless, capable junior analyst?" Prioritise processes that are rules-based, digitally native, and high-volume.

2. Prototype with 'human-in-the-loop' agents. The first step toward autonomy is not to remove humans, but to augment them. Start designing workflows where an AI carries out 80% of the work, but a human staff member is required to review, approve, and direct at critical checkpoints. This builds institutional trust and helps mitigate risks while you learn the technology's capabilities and limitations.

3. Invest in new skills. The skill of 'prompt engineering' is evolving into 'AI direction' or 'workflow orchestration'. Train your team to think in terms of defining clear goals, providing necessary resources (documents, API access), and setting explicit constraints and success criteria for an AI. This is a management skill, not just a technical one.

4. Evaluate the vendor landscape. Anthropic's Fable 5 is a major signal, but it won't be the only option. Monitor announcements from OpenAI, Google, and a growing ecosystem of AI agent startups. When evaluating platforms, pay close attention to their frameworks for safety, reliability, monitoring, and debugging. The best agentic platform for your business will be the one that aligns with your specific risk tolerance and operational needs.