Blog Details

The clearest warning signs when evaluating an AI vendor in Singapore are: guaranteed results or grant approval, no clear pricing or scope, pressure to sign quickly, no verifiable track record, and vague answers about data handling. Any one of these alone might have an innocent explanation. Two or more together is reason to walk away or dig much deeper before signing.

Singapore's AI consulting space has grown fast, and with growth comes a mix of genuinely capable operators and people who learned the buzzwords faster than they learned the delivery. Here are the 9 red flags worth checking for before you commit budget.

1. Guaranteed results before any assessment

No credible consultant can promise a specific ROI, accuracy rate, or business outcome before they've looked at your data and systems. "We guarantee 30% cost savings" without having seen your operations is a sales line, not an assessment. Real numbers come after a proper readiness audit, and even then they're estimates, not guarantees.

2. Guaranteed grant approval

This is one of the clearest scam signals in the Singapore market specifically. No consultant, agency, or grant advisor can guarantee EDG, PSG, or any other grant will be approved. Approval is at the granting agency's discretion, requires pre-approval before work starts, and the business owner submits the application in their own name. Anyone promising "guaranteed approval" or "we'll get your grant through" is either misrepresenting how the process works or planning to cut corners on the application.

3. No written scope or deliverables

If a vendor won't put in writing exactly what you're getting (specific deliverables, timeline, what's in and out of scope), you have no way to hold them accountable later. Verbal promises during a sales call mean nothing once the invoice is paid.

4. Pressure to sign immediately

"This price is only valid today" or "we can only take you on if you sign this week" is a pressure tactic, not a genuine capacity constraint in most cases. Legitimate vendors understand that a S$20,000+ decision deserves proper consideration time.

5. No verifiable track record

Ask for specific examples of similar work and, ideally, a reference you can actually speak to. A vendor who can only offer vague claims ("we've helped many businesses") or stock case studies with no verifiable details is a risk, especially for a first-time buyer who doesn't yet know what good delivery looks like.

6. Vague or evasive answers about data handling

Ask directly where your data will be stored, who has access, and whether it's used to train any third-party AI model. A vendor who can't answer clearly, or brushes off the question, either hasn't thought it through or doesn't want to commit to an answer. This has real PDPA implications; see PDPA compliance for AI chatbots.

7. Pricing that seems too good to be true

A custom chatbot build quoted at S$1,500 when the market range is S$8,000+ usually means one of three things: it's a bare template with minimal real customisation, there will be change orders once you're committed, or the vendor is underpricing to win the deal and won't be able to sustain proper support afterward. See realistic ranges in our AI chatbot pricing guide.

8. No post-launch support offered

If a vendor's plan ends the moment the system goes live, with zero support window, you're on your own the first time something breaks or behaves unexpectedly, which is common in the first weeks of real usage. A responsible vendor builds in at least 30 days of post-launch support by default.

9. Can't explain their own recommendation in plain language

If you ask "why this approach and not a simpler one" and get buzzword soup instead of a clear, specific answer, that's a sign the vendor may be pattern-matching a generic pitch to your business rather than actually thinking through your situation. A good consultant can explain their reasoning to a non-technical business owner.

What to do if you spot a red flag

• Ask the direct follow-up question. Sometimes there's a reasonable explanation, and how they respond tells you a lot.
• Get everything in writing before proceeding further, including anything promised verbally.
• Get a second quote or opinion from another vendor to compare, especially if something feels off but you can't quite articulate why.
• Check ACRA for business registration and look for any independently verifiable reviews or references.

What good vendors do instead

The inverse of each red flag is a reasonable checklist: realistic, assessment-based estimates rather than guarantees; honest grant language ("we'll help you prepare a strong application, but approval is never guaranteed"); written scope documents; no pressure tactics; verifiable past work; clear data handling policies; market-realistic pricing; a defined support window; and plain-language explanations of their recommendations. Run through our full 12 questions to ask when choosing an AI consultant to test for these directly.

Red flags specific to grant-linked pitches

Because so much of the SG AI vendor market is now tied to grant funding, a specific pattern of red flag has become common: vendors who lead the entire sales conversation with the grant percentage rather than the actual solution. If a pitch spends more time on "you'll only pay 30% out of pocket" than on what's actually being built and why it fits your business, that's worth pausing on. The grant should support a project you'd want regardless of funding, not be the reason the project exists in the first place. Similarly, be cautious of any vendor who suggests inflating quoted costs to "maximise the grant amount," which is not how legitimate grant claims work and carries real compliance risk for you as the applicant.

How to handle it if you've already paid a vendor showing red flags

If you're partway into a project and starting to notice these signs, don't wait for things to fully unravel before acting. Request a written status update and a revised timeline in writing. If answers remain vague, consider bringing in an independent second opinion, even a paid one-off consultation with another firm, to assess what's actually been delivered against what was promised. For grant-funded projects specifically, keep your own documentation (emails, invoices, meeting notes) thorough, since this matters both for any dispute resolution and for your own grant claim compliance.

A note on Freemansland's own approach

We won't guarantee your grant will be approved, and we won't promise a specific ROI before we've looked at your business. What we will do is give you an honest, assessment-based view, in writing, before you commit any budget. If that's a lower bar than what another vendor is promising you, that's worth noticing.

How these red flags show up differently online vs. in person

A polished website and confident LinkedIn presence are not proof of legitimacy on their own; these are relatively easy to build regardless of actual delivery capability. Conversely, a vendor with a modest website but strong, verifiable references and clear, specific answers in conversation is often the safer bet. Weight what you learn from a direct conversation more heavily than marketing polish, and treat glossy case studies with generic, unverifiable numbers ("increased efficiency by 40%" with no context on how that was measured) with the same skepticism as a guaranteed-ROI sales pitch.

A short pre-signing checklist worth running through once

Before you sign anything, it's worth a final pass through a short checklist rather than relying purely on gut feel from the sales conversation: is the scope in writing, with specific deliverables rather than vague descriptions? Is the price clear, including what's excluded? Is there a defined support period after launch? Have you spoken to at least one verifiable reference or seen a working example of similar past work? Is the data handling approach clear and PDPA-compliant? If you can answer yes to all five, you've covered most of the ground the red flags above are designed to catch.

What a healthy vendor relationship looks like once you're past the red flags

Once you've selected a vendor and started working together, healthy signs continue beyond the initial sales process: regular, proactive updates rather than only responding when you chase them, willingness to say "we're not sure yet" rather than guessing confidently on technical questions, and comfort with you asking to see work in progress rather than only a finished demo. If the vendor's behaviour changes noticeably for the worse once the contract is signed and the deposit is paid, that shift itself is worth taking seriously and raising directly rather than assuming it will improve on its own.

Ready to see what AI can do for your business?

If you're currently comparing vendors and something feels off, we're happy to give you a second opinion, no obligation. Learn about our AI strategy and advisory approach or browse all services. Request a quote or get in touch. WhatsApp +65 9184 9908 or email glenn@freemansland.co.

Frequently Asked Questions