Why This Matters to Your Business

You may never need to manage a class action lawsuit, but you do face a similar business risk every day: promising a feature, service level, delivery date or business outcome before your team can reliably deliver it.

That risk becomes more serious when your customers see the promise repeatedly through advertisements, sales presentations, websites or social media. Apple’s Siri settlement is a useful reminder that expectations are shaped by what you say publicly, not only by what your product can do today.

Apple agreed to a reported $250 million settlement in a United States class action lawsuit concerning delayed Siri AI features. Eligible customers may receive up to $95 per claim, although the final amount depends on the number of valid claims.

TL;DR

For your SME, the lesson is simple: make sure your marketing promises match what your team can deliver now.

Document product limitations, label planned features clearly, and create a review process before publishing claims about automation, artificial intelligence or service performance.

What This Means

Apple previewed personalised Siri capabilities in June 2024, including understanding a user’s personal context, recognising what was on the screen and taking deeper actions across apps. One example showed Siri using information from Mail and Messages to answer questions about a family member’s flight and lunch reservation.

Those capabilities were presented publicly, including through product presentations, Apple’s website and a television commercial. Apple later delayed the personalised Siri launch, and a lawsuit alleged that the company’s advertising created misleading expectations. Apple agreed to settle while stating that it wanted to stay focused on delivering its products and services.

According to the reported proposed timeline, eligible United States customers would likely start receiving notifications and submitting claims around 21 September 2026. The proposed claims period would end on 21 December 2026, giving customers approximately 90 days to respond. A final approval hearing is scheduled for 24 February 2027.

This case applies only to eligible customers in the United States who purchased specified iPhone models during the stated period. It is not a Malaysian consumer claim process, and Malaysian iPhone owners should not assume they qualify.

The practical lesson: customers judge your business by the gap between what you promised and what they actually experienced.

How This Applies to Malaysian SMEs

1. Be precise when selling automation. Suppose you advertise that a WhatsApp automation system will “handle all customer enquiries automatically”. In practice, it may answer common questions but still need a staff member for complaints, unusual requests, payment issues or Bahasa Malaysia slang. If your wording suggests full automation, customers may reasonably expect no human involvement. A better statement would explain the scope: the system handles common questions, captures customer details and routes complex cases to your team.

This matters for Malaysian businesses because customers often contact you outside office hours through WhatsApp, Instagram, Facebook and Google Business Profile. A message such as “instant response” can create a clear expectation. If the system only replies after a delay, or gives a generic message without recording the enquiry, your customer may feel misled even if the software is functioning as designed.

2. Separate available features from planned features. A renovation company may tell prospects that its new customer portal will show live project progress, approve material selections and provide automatic updates. If those functions are still being built, describe them as planned or in development. Do not present a future workflow as though it is already available.

The same principle applies to accounting, inventory and appointment systems. You might plan to connect your point-of-sale system to stock management, but if the connection has not been tested, avoid promising real-time inventory accuracy. Instead, identify what is working, what requires manual checking and what will be added later.

3. Keep evidence for every important claim. If your website says you reply within 10 minutes, keep records showing how that figure was calculated. If you claim that an automation reduces manual data entry, document the process before and after implementation. If you advertise “24-hour support”, define whether that means a human response, an automated acknowledgement or emergency support.

These records help you improve operations and answer customer questions. They also stop marketing language from becoming detached from the reality of your staff capacity. For a small business with one to 50 employees, a shared folder containing approved claims, test results, service definitions and customer-facing explanations can be enough to create useful control.

4. Treat AI output as assistance, not guaranteed judgement. AI tools can draft replies, summarise enquiries and classify leads, but they may misunderstand mixed-language messages, local place names, product variations or sensitive complaints. If your business allows AI to respond directly, define which topics require human approval.

For example, a clinic may allow automation to answer questions about opening hours and appointment availability, but require staff review for symptoms and treatment advice. A logistics company may automate delivery-status updates but route damaged-goods complaints to a trained employee. Clear boundaries reduce the chance that a confident automated answer becomes an operational or reputational problem.

Common Promise Gaps to Check

Customer-facing promise	What to verify internally	Safer wording
“Instant reply”	Response time, channel coverage and failure alerts	“Automatic acknowledgement when your message is received”
“Fully automated”	Exceptions, approvals and human handovers	“Automates routine steps and sends complex cases to staff”
“Real-time stock”	Sync frequency, manual adjustments and offline sales	“Stock information is updated through the connected system”
“Available 24/7”	Whether support is automated or handled by people	“Submit enquiries anytime; team replies during stated hours”
“AI understands your business”	Data sources, permissions and accuracy checks	“AI assists with selected tasks using approved business information”

Practical Takeaways

• List every public claim about speed, accuracy, automation, availability and results.
• Mark each claim as available now, limited, in testing or planned.
• Ask the person responsible for operations to approve marketing claims before publication.
• Explain clearly when a customer will interact with a chatbot, workflow or human employee.
• Set an escalation path for complaints, unusual requests and sensitive information.
• Test automated replies in English, Bahasa Malaysia and the mixed language your customers actually use.
• Review claims whenever software, staff capacity, suppliers or operating procedures change.
• Keep screenshots, process notes and performance records for important service statements.
• Train staff to correct an inaccurate promise quickly rather than defending it.

A Simple Review Process for Your Team

Before publishing a new claim, ask four questions. First, what exactly will the customer expect? Second, can the current system deliver that result consistently? Third, what happens when the normal process fails? Fourth, who is responsible for correcting the customer and improving the process?

You can run this review in a short weekly meeting. Select one website page, advertisement, brochure or automated message. Compare it with the actual customer journey. If the wording is broader than the capability, narrow the wording or fix the process before promoting it further.

This approach is especially useful when adopting new software. Instead of asking only whether a tool has an impressive feature, ask whether your business has the data, approval rules, staff roles and fallback process needed to use it safely.

The Bigger Picture

The Siri dispute shows how technology announcements can create expectations long before the underlying feature is ready. The same pattern can affect smaller businesses when new automation tools are marketed as if setup alone will solve customer service, sales or administrative problems.

Customers are becoming more familiar with chatbots, AI assistants and automated workflows. That makes honest explanations more valuable, not less. When you clearly describe what automation can do, where it stops and when a person will step in, customers can use the system with realistic expectations.

The long-term advantage is operational trust. Your business does not need to promise futuristic capabilities. It needs to deliver a dependable experience, measure the parts that matter and communicate changes early. Start by reviewing your top five customer-facing promises this week. If your team cannot demonstrate each one, revise the wording or improve the workflow before making the promise louder.

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 →