OpenAI CEO Sam Altman said the company does not expect to pursue an initial public offering until it can make stronger assurances about the safety of its AI models. The remarks, reported after OpenAI's DevDay event, do not set a timeline for any IPO, but they highlight how closely business strategy, regulation and AI risk management are becoming linked.

For European businesses using or evaluating generative AI, the immediate impact is not about a stock market listing. The more important signal is that leading AI providers are publicly framing model safety as a prerequisite for long-term growth and governance.

What happened

According to The Verge, Altman said OpenAI intends to continue advancing its models, but does not want to go public before the company can make more confident safety claims. He did not give a firm schedule for any listing.

These comments come amid broader scrutiny of advanced AI systems, including questions about how providers test models, manage misuse risks and communicate limitations to customers, developers and regulators.

Why it matters for European businesses

The statement is relevant because it shows that AI safety is no longer only a research or policy topic. It is increasingly tied to corporate governance, investor expectations, customer trust and future regulation.

For companies in Europe, especially those adopting large language models for customer service, content generation, internal automation or software workflows, this matters in several ways:

• Vendor risk is becoming more important. Businesses relying on third-party AI tools need to assess not just features and pricing, but also how providers handle testing, model limitations, misuse prevention and incident response.
• AI procurement is becoming more compliance-driven. Enterprises and regulated sectors are likely to ask for clearer documentation, safety commitments and governance processes from AI vendors.
• Trust affects deployment. If AI providers cannot give clear assurances on reliability and safety, some businesses may slow down deployments in sensitive use cases.
• Regulatory pressure in Europe remains relevant. Even when a story concerns a US AI company, European rules and buyer expectations can shape how AI products are adopted, documented and integrated into business processes.

Who may be affected

The development is most relevant for:

• SMEs adopting generative AI for marketing, support, operations or internal productivity.
• IT and digital teams responsible for selecting AI tools and integrating them into websites, CRMs, knowledge bases or workflows.
• Regulated businesses that need stronger assurances before using AI in higher-risk or customer-facing contexts.
• Agencies and software providers building services on top of major foundation model platforms.

What companies should consider

European businesses do not need to react to this as a direct legal or operational change, but the comments are a useful reminder to strengthen AI governance before dependence on external models grows further.

• Review AI suppliers. Ask what safety testing, monitoring and documentation your provider offers, especially for business-critical use cases.
• Match AI use to risk level. Low-risk tasks such as drafting or summarisation may need lighter controls than customer advice, decision support or regulated workflows.
• Keep human oversight where needed. For external communications, compliance-sensitive outputs and operational decisions, human review remains important.
• Document limitations internally. Teams should understand where AI tools can produce errors, biased outputs or unsupported claims.
• Plan for vendor changes. Businesses building on external AI platforms should reduce overdependence by keeping clear processes, fallback options and data governance controls.

Altman's comments do not create a new rule for AI buyers, and they do not confirm any future IPO timetable. But they do reinforce a broader market reality: as AI systems become more capable, safety and governance are moving closer to the centre of commercial decision-making.