Why leadership changes matter to your AI plans
You may not be running a global AI company, but you still depend on people to make technology work. If your business uses automation for enquiries, sales follow-ups, customer support, documents, or internal reporting, a change in the person responsible can affect how smoothly those systems operate.
The recent departure of Chris Malone, OpenAI’s former head of data centres, is notable because he was responsible for a major infrastructure function. OpenAI said it had reorganised its infrastructure organisation and still had an experienced team in place, while reporting also pointed to several leaders sharing responsibility for data centre strategy. TechCrunch
For you as a Malaysian SME owner, the lesson is not to speculate about one executive’s departure. It is to ask a more useful question: Can your business continue operating if one key person, software provider, or technology partner changes direction?
TL;DR
AI projects should not depend on one person’s memory, one vendor’s promises, or one undocumented process.
Build clear ownership, keep your business data organised, document workflows, and review AI suppliers before you make the technology part of daily operations.
What This Means
Infrastructure is the foundation that allows digital services to run. For a large AI company, that can include data centres, computing capacity, energy planning, engineering teams, delivery programmes, and technical leadership. When responsibility is reorganised, the company needs clear handovers and strong processes so projects do not slow down.
OpenAI’s reported executive turnover gives us a practical reminder about organisational resilience. TechCrunch reported that the company had recorded 13 executive departures in 2026, including senior leaders in revenue, operations, product, ethics, marketing, and infrastructure. TechCrunch
High-profile departures do not automatically mean that a company’s products will fail. Large organisations often have several people sharing responsibilities, and leadership changes may be part of restructuring. However, frequent changes can create uncertainty about priorities, product direction, support quality, security practices, and accountability.
That same principle applies to a smaller company. If only one employee knows how your automated quotation process works, or only one external consultant can update your chatbot, your business has a hidden operational risk. The system may appear efficient, but it is fragile underneath.
Good automation is not just about using advanced software. It is about making sure your business can keep working when people, vendors, or priorities change.
How This Applies to Malaysian SMEs
1. Do not let one employee become the entire system. Imagine your sales administrator has created an automated process that captures leads from WhatsApp, assigns them to sales staff, sends follow-up reminders, and records the outcome in a spreadsheet or CRM. If that employee resigns, takes extended leave, or moves to another role, the process may stop because nobody else knows the settings, passwords, logic, or exception rules.
You can reduce this risk by documenting each workflow in plain language. Record what starts the process, what information is required, what happens next, who checks the result, and what to do when something goes wrong. Keep a second person familiar with the workflow. This does not require technical knowledge; it requires a simple operating guide that another responsible employee can follow.
2. Treat your technology provider as a business dependency. Many Malaysian SMEs work with software vendors, automation agencies, accountants, marketing platforms, payment providers, or cloud applications. If a provider changes its product, support team, ownership, or service terms, your daily processes may be affected. A provider does not need to shut down for disruption to happen. A key feature may be removed, an integration may stop working, or support may become slower.
Before adopting a system, ask who owns your data, how you can export it, who can access the account, and what happens if you need to change providers. Keep administrator access under your company’s control rather than allowing it to remain with a former employee or external consultant. You should also know which manual process can keep essential work moving during an outage.
3. Separate useful automation from important business judgement. A café chain might use automation to collect catering enquiries and send standard replies. A renovation company might use it to organise site visit requests. A wholesaler might use it to remind customers about repeat orders. These are useful tasks, but the system should not make every decision without review.
For example, an automated reply can acknowledge a customer’s message, but a person should confirm unusual requests, complaints, refunds, credit terms, or changes to delivery commitments. In Malaysia, businesses may also handle information involving names, phone numbers, addresses, identification details, or payment records. Your workflow should define which information can be processed automatically and which cases must be escalated.
4. Plan for changing AI capabilities. AI tools are developing quickly, and suppliers may change models, features, limits, or recommended uses. A workflow that performs well today may produce different results after a product update. You should therefore test important outputs rather than assuming that an automated result is always correct.
If you use AI to draft customer messages, summarise documents, classify enquiries, or prepare reports, define a quality check. Decide who reviews the output, what errors matter most, and how corrections are recorded. This is especially important when your reputation depends on accurate information, such as product specifications, quotations, delivery dates, or regulatory documents.
A simple risk snapshot for your business
| Business dependency | Possible disruption | Practical safeguard |
|---|---|---|
| One employee understands the workflow | Process stops when that person is unavailable | Create a step-by-step guide and train a backup |
| One software provider handles key operations | Feature, support, or access changes | Keep admin ownership and export essential records |
| AI-generated customer replies | Incorrect or unsuitable information is sent | Set approval rules for sensitive messages |
| Automated lead or order records | Missing data creates follow-up failures | Use required fields and weekly exception checks |
| Undocumented passwords and integrations | Recovery takes too long after staff changes | Maintain a secure access register and recovery contact |
The table is a management checklist rather than a prediction about any particular supplier. Its purpose is to help you identify where your business depends on a single point of failure.
Practical Takeaways
- List your essential automated workflows. Include sales enquiries, customer support, purchasing, payroll administration, inventory updates, invoicing, and reporting.
- Name an owner and a backup. Every important workflow should have at least two people who understand its purpose and basic operation.
- Document the process. Write the trigger, steps, approvals, exceptions, and final result in language your staff can understand.
- Control administrator access. Use company-managed accounts and remove access promptly when staff or contractors leave.
- Keep an exportable record. Make sure important customer, supplier, and operational information is not trapped in one system.
- Define human approval points. Require review for complaints, refunds, unusual orders, sensitive personal information, and legally important communications.
- Test recovery. Ask whether your team could continue essential work if a system became unavailable for one working day.
- Review suppliers regularly. Check product changes, support contacts, security information, access rights, and your ability to move data.
- Measure outcomes, not excitement. Track response times, missed follow-ups, correction rates, and staff adoption to see whether automation is helping.
The Bigger Picture
The AI industry is being built at a remarkable pace. OpenAI’s Stargate project was described as a $500 million data centre initiative involving OpenAI, Oracle, Nvidia, SoftBank, and Microsoft. TechCrunch Large projects of this scale require major infrastructure, specialist teams, and clear coordination. They also show why technology alone is never the complete story.
For SMEs, the long-term direction is straightforward: more business tools will include AI features, but dependable results will still rely on good processes, clean information, trained staff, and sensible oversight. You do not need to copy the structure of a large AI company. You do need the same basic discipline of knowing who is responsible, how work moves, and what happens when something changes.
Reported concerns about OpenAI’s leadership churn, safety teams, valuation, and future public listing are matters for the company and its stakeholders. TechCrunch reported that an expected public listing had reportedly shifted from 2026 to 2027. TechCrunch For your company, the useful takeaway is more immediate: do not build critical operations around assumptions you have not checked.
Start with one workflow that causes repeated delays, such as responding to enquiries or preparing quotations. Document it, assign ownership, add a review step, and make sure someone else can operate it. Once that foundation is sound, you can expand carefully into other areas.
The best AI setup for your SME is not the one with the most impressive features. It is the one your team understands, can supervise, and can recover when people or platforms change.
Ready to Streamline Your Operations?
Your business should run itself. AutoRunBiz deploys AI agents to automate your daily operations — WhatsApp orders, invoicing, customer follow-ups, and accounting. Book a free 15-min ops audit to see where automation fits your business →