AI Is Moving from Experiment to Management Responsibility
Artificial intelligence is no longer discussed only by engineers and technology companies. It is now part of public policy, workplace planning, online safety, surveillance, data infrastructure and education. In a recent The Verge interview, New York Governor Kathy Hochul argued that AI should be “less evil” while discussing job displacement, data centres, children’s online safety and the responsibilities of government.
For you as a Malaysian SME owner, the important lesson is not what happens in New York alone. The bigger issue is that expectations around responsible technology are rising. Your customers, employees and business partners may soon ask how your company uses AI, what information you provide to AI tools, and whether a human checks important decisions. Waiting until a mistake becomes public is a poor way to create an AI policy.
What Happened
The interview covered several technology policy issues. Hochul discussed online restrictions intended to protect teenagers, including rules affecting how young people use platforms such as Instagram. The article noted that broad age verification could also require adults to show identification online, raising questions about privacy and anonymity. These concerns show how a rule designed for safety can create new obligations and unintended consequences for ordinary users, companies and platforms.
Hochul also spoke about New York’s restrictions on “ghost guns”, including laws intended to prevent 3D printers from producing gun parts. The interview explored whether technical workarounds could weaken such restrictions. This is a useful reminder that technology regulation is not simply about writing a rule. Policymakers must understand how systems actually work, how users may adapt to restrictions, and how enforcement can be carried out in practice.
Another major issue was the backlash against AI data centres. According to The Verge, Hochul signed a one-year moratorium on data centre construction in July and was asked what would make future data centre development acceptable. The discussion also covered AI-related job displacement and the need to train workers for changing roles. Hochul described bringing in experts to examine which jobs are most vulnerable and how people could prepare for new opportunities.
“Bring in all the information as quickly as possible, pressure-test it, and then go with my gut,” Hochul said when describing her decision-making framework, as reported by The Verge.
Why This Matters for Malaysian SMEs
Malaysian SMEs are already using AI in practical ways: drafting customer replies, summarising meetings, creating product descriptions, translating messages, preparing social media content and analysing sales records. These uses can improve productivity, but they also introduce risks. An employee may paste a customer’s personal information into a public AI chatbot. A generated reply may contain an incorrect warranty promise. An AI-written marketing message may make a claim your business cannot support.
These are not abstract technology concerns. A Malaysian retailer could use AI to recommend products based on purchase history. A workshop could use it to draft service reminders. A tuition centre could use it to prepare learning materials. A small professional firm could use it to summarise documents. In each case, you need to decide what information may be entered, who reviews the result and how long records are retained.
The New York debate also highlights a second issue: safety rules can affect everyone, not only the people they are designed to protect. If your SME introduces strict access controls, automated monitoring or identity checks, customers may find the process inconvenient or intrusive. You therefore need to balance protection with user trust. Collecting more data is not automatically safer. It can create a larger target if accounts, devices or databases are compromised.
AI also changes job design. This does not necessarily mean replacing employees. For a small Malaysian business, the more useful approach is often to redesign routine work so employees can focus on customer service, quality checks, sales conversations and problem-solving. For example, an AI tool can prepare a first draft of a quotation, while your staff verifies technical details, delivery commitments and terms.
Practical AI Controls for Your Business
| Business area | Useful AI role | Control you should apply |
|---|---|---|
| Customer service | Draft replies and classify enquiries | Human approval for complaints, refunds and legal matters |
| Marketing | Generate campaign ideas and content drafts | Check claims, tone, trademarks and local context |
| Finance administration | Extract information from documents | Verify figures before payment, invoicing or reporting |
| Human resources | Prepare interview questions and training materials | Do not let AI make final hiring or disciplinary decisions |
| Operations | Summarise work orders and identify recurring issues | Restrict access to confidential customer and supplier data |
What “Less Evil” Should Mean in Your Company
“Less evil” is not a technical standard, but it is a useful management question. Before adopting an AI tool, ask whether it creates more value than risk for the people affected by it. Does it make a process clearer or more confusing? Does it help employees or quietly remove their ability to challenge a decision? Does it use information that customers reasonably expect you to protect?
You can turn these questions into a short internal policy. State that employees must not enter passwords, identity documents, bank details, confidential contracts or unnecessary personal information into unapproved AI services. Require a human review before sending externally visible content. Keep a simple record of which AI tools are used, for what purpose and by which team. Review the list every few months as tools and business needs change.
For higher-risk tasks, add an escalation rule. AI should not independently approve loans, reject applicants, diagnose health conditions, make disciplinary findings or provide legal conclusions. A person with appropriate authority should review the recommendation and be accountable for the final decision. This approach is especially important when your company serves children, older customers or people sharing sensitive information.
The Bigger Picture
The New York discussion shows that AI policy is becoming connected to infrastructure, employment, privacy and public trust. Data centres require physical resources and affect surrounding communities. Automated systems can change how work is organised. Safety rules can influence how people access online services. Technology decisions therefore cannot be left only to the person who first installs an app or signs up for a chatbot.
For Malaysian SMEs, this does not mean creating a large compliance department. It means assigning ownership. One person should maintain your approved AI tool list, document basic risks and coordinate staff training. Your policy can be short, written in plain language and tailored to your actual workflows. The key is to make responsible use part of daily operations rather than a document that nobody reads.
Start with one low-risk process, such as drafting frequently asked questions or summarising internal meeting notes. Measure whether the result saves time without reducing accuracy. Then add review steps before expanding to customer data or operational decisions. In this way, you can gain the benefits of automation while keeping human judgement where it matters most.
The strongest lesson from Hochul’s remarks is simple: adopt technology with ambition, but govern it with discipline. AI can help your business respond faster and work more consistently. Your responsibility is to ensure that faster does not mean careless, and that automation does not remove accountability.
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 →