Three signals that prove AI has crossed from experiment to infrastructure.

Signal 1 — Anthropic files confidential S-1 with the SEC

On 1 June 2026, Anthropic submitted a draft Form S-1 registration statement to the US Securities and Exchange Commission, confidentially, for a proposed IPO of its common stock. The company currently runs at a $47 billion annual revenue rate — up from $10 billion just twelve months ago — and carries a $965 billion valuation from its May Series H raise.

For a business operator, the signal here is not about buying shares. It is about what institutional capital at this scale tells you about permanence. Sovereign wealth funds, Google, and Amazon do not back a company at near-$1 trillion because they think the technology might be interesting. They back it because the revenue growth is real, the moat is wide, and the demand is structural.

If you have been waiting for AI to "stabilise before committing", the companies whose tools you would use are now entering the public markets. That is a different kind of signal than a product update.

Signal 2 — OpenAI replaces Custom GPTs with Workspace Agents

OpenAI announced Workspace Agents as the successor to Custom GPTs — shared AI agents that run inside ChatGPT Business and Enterprise accounts, connect to third-party apps, and handle repeatable multi-step tasks without supervision. The agent integrations include Slack, Google Drive, Microsoft 365, Salesforce, Notion, and Atlassian Rovo.

The free preview runs until 6 July 2026. After that, credit-based pricing begins. For a small business that has been meaning to automate customer follow-up, proposal drafts, or inbox routing — this is a concrete window with a specific close date. A sales opportunity agent, for instance, can research accounts, summarise call notes, and post deal briefs directly into Slack without a human triggering any of it.

The shift from Custom GPTs (which you still had to prompt manually) to Workspace Agents (which run autonomously on triggers and schedules) is the difference between a tool you use and a system that runs your workflow.

Signal 3 — 54% of UK SMEs now use AI, up from 25% in 2024

The British Chambers of Commerce published data showing that 54% of UK SMEs are now using AI tools — more than double the 25% recorded in 2024. The same data shows that roughly one in five firms with bespoke AI integration reports staffing reductions attributable to AI, and firms using AI are approximately three times more likely to have restructured job roles.

Entry-level roles are under particular pressure. Graduate hiring is down 45% year on year among AI-adopting firms. Youth unemployment is forecast to reach 17% this year partly because tasks traditionally done by junior staff — document drafting, basic analysis, customer interaction — are increasingly handled by AI.

For a service business owner, this data has two readings. First: more than half your competitor set is already using AI to reduce costs, accelerate delivery, and reduce headcount risk. Second: the businesses leading on AI adoption are pulling away from those that are watching and waiting. The cost of not acting has risen.