Why an AI Mathematics Story Should Matter to Your Business
A story about OpenAI attempting to solve the Navier–Stokes problem may sound far removed from your daily business concerns. You may be running a workshop in Shah Alam, a trading company in Penang, a clinic in Johor, or a small professional-services firm in Kuala Lumpur. Yet this story points to a practical issue you cannot ignore: AI is becoming powerful enough to work on difficult, specialised problems, while the way companies collect, use and protect information is still being debated.
According to The Verge, OpenAI said an unreleased model used roughly 10,000 AI agents and found a solution to the Navier–Stokes problem in 88 hours. The problem concerns the movement of fluids and has challenged mathematicians for close to 90 years, according to the same report. The announcement created excitement, but also concern over research ownership, competition and whether AI systems may have been indirectly influenced by users’ work.
For your company, the lesson is not that you need thousands of AI agents. The lesson is that every AI tool you use may sit inside a wider system involving training data, conversation logs, access permissions and unclear ownership. If you allow employees to paste customer records, technical designs, supplier terms or internal plans into an AI service, you need to know what happens next.
What Happened
OpenAI said its model worked on the famous Navier–Stokes problem after hearing that other researchers might be making progress on Millennium Prize problems. The company reportedly assembled a large swarm of AI agents to investigate the problem and described the result as a milestone in AI-assisted mathematics. The Clay Mathematics Institute, which administers the prize, had not awarded the bounty at the time of the report.
The timing caused controversy. One day before OpenAI’s announcement, New York University mathematics professor Tristan Buckmaster published findings on a related problem with Levent Alpöge, a researcher associated with Anthropic, according to The Verge. Buckmaster alleged that OpenAI knew about their progress and questioned whether his work had been accessed through Codex sessions. OpenAI denied accessing specific user data, but said it could not completely rule out the possibility that de-identified data derived from product usage had helped improve its models.
The report also described a dispute over research credit and communications between Buckmaster and an OpenAI researcher. The exact sequence of events remains contested. That uncertainty is important: when AI produces an answer, it can be difficult to identify precisely which information influenced the result, especially when training data, public material, user interactions and model-generated ideas overlap.
“Mathematics depends heavily on an informal norm of trust,” said Matthew Ballard, a mathematics professor quoted by The Verge. For a business, the equivalent is simple: your employees, customers and partners must be able to trust how information is handled.
Why This Matters for Malaysian SMEs
Many Malaysian SMEs are adopting AI through everyday tools rather than large technology projects. You may use an AI assistant to draft quotations, summarise meetings, translate product descriptions, answer customer enquiries or prepare social media content. Your staff may also use coding assistants, design platforms and automated customer-service tools without informing you.
That convenience creates information risks. A renovation company might upload a client’s floor plan to obtain a project summary. A food manufacturer might paste a formulation into an AI assistant for label suggestions. An accounting practice might ask an AI tool to classify transactions using copied invoices. A recruitment agency might upload candidate profiles to create interview questions. In each case, the information could include personal data, confidential commercial details or intellectual property.
Malaysia’s Personal Data Protection Act 2010 applies to the processing of personal data in commercial transactions, and the Personal Data Protection Commissioner provides guidance and regulatory information. You should therefore treat AI use as part of your existing data-governance responsibilities, not as an informal experiment. If personal information leaves your approved systems, you need a clear reason, appropriate safeguards and a way to respond if something goes wrong.
The story also matters because AI can produce impressive results without providing a complete audit trail. An employee may receive a highly convincing answer but be unable to explain which source material influenced it. That is dangerous when the output affects safety, contracts, financial reporting, medical information or engineering decisions. A small company does not need to reject AI, but it does need to place human review around high-impact work.
Practical AI safeguards for your business
| Risk | What you should do |
|---|---|
| Confidential information pasted into public tools | Create a rule that customer, supplier and product secrets stay out of unapproved AI services. |
| Personal data uploaded without review | Remove names, identification numbers, addresses and contact details before using AI. |
| Unverified AI-generated advice | Require a staff member with relevant knowledge to check important outputs. |
| Unclear ownership of generated work | Record which tool was used, who reviewed the output and which original materials were supplied. |
| Employees using tools secretly | Provide an approved list of tools and explain acceptable use in plain language. |
The Bigger Picture
The OpenAI mathematics episode shows that AI competition is moving quickly. Companies can mobilise large computing systems to investigate complex questions in a short period. The speed is attractive for SMEs: a tool may help you compare suppliers, identify recurring customer complaints, test marketing ideas or organise operational data far faster than manual work.
However, speed can expose weaknesses in business processes. If you do not know what data enters an AI tool, who can access it, how long it is retained or whether it may be used to improve the service, you are not fully in control. The concern raised by mathematicians about private research being exposed has a direct SME equivalent: a competitor should not gain an advantage from your product roadmap, customer list or technical know-how.
Start with a one-page AI policy. State which tools employees may use, what information is prohibited, when manager approval is required and which tasks always need human checking. Keep an AI-use register containing the tool name, purpose, department and type of data involved. Review the register monthly, especially when staff introduce new browser extensions or workplace applications.
You should also separate low-risk and high-risk uses. Drafting a generic festive greeting is low risk. Preparing a response based on a customer’s medical information, creating legal contract language or recommending an engineering specification is high risk. For the latter category, use AI only as an assistant and require qualified human review before anything is sent, approved or implemented.
Finally, ask vendors direct questions. Where is your data stored? Is it used to train models? Can training be disabled? How long are prompts retained? Who can access them? Can your organisation delete the information? Keep the answers in writing. If a vendor cannot explain its data practices clearly, do not give it your most sensitive information.
The future of AI in Malaysian SMEs will not be decided only by spectacular breakthroughs in mathematics. It will be decided by whether owners like you can combine useful automation with disciplined information handling. Adopt AI where it removes repetitive work, but protect the knowledge that makes your business competitive. Trust, transparency and review are not obstacles to automation; they are what allow you to use it confidently.
Ready to Streamline Your Operations?
Technology moves fast. Your operations should keep up. AutoRunBiz builds AI systems that run your daily workflows — from WhatsApp order capture to accounting. Book a free 15-min ops audit →