A Malaysian AI vendor announced a deepened Google Cloud partnership and a 30 fold efficiency claim. Six weeks in, no automaker, no third party benchmark, and no audited conversion number has surfaced to back it up.
At a roundtable inside Google's Asia-Pacific headquarters in Singapore last month, a Malaysian AI vendor called Aimotion announced it had cracked what the automotive industry has spent years chasing: a 30-fold gain in marketing efficiency. Six weeks later, the company cannot point to a single automaker, dealer group, or named customer willing to go on record as proof.
The May 23 event, branded "Boundless Cloud, Global Motion," was the public stage for a deepened collaboration between Aimotion and Google Cloud, announced via PR Newswire. On the marquee metric, Aimotion says its video production workflow now runs in 10 minutes, down from four hours. That is the source of the "over 30x" line that has followed the announcement. It is also the figure with the thinnest public backing.
The press release carries three other headline numbers, all Aimotion-self-reported: a 21% drop in factual errors and hallucinations versus general-purpose models, a content output rate above 93% meeting internal quality thresholds, and cross-lingual content checks measured in milliseconds. BigGo Finance, an independent aggregator, restates the same metric table and explicitly labels the figures as "Aimotion's reported internal benchmarks," not third-party audited numbers. That distinction is the spine of the story.
The partnership is real. Aimotion describes its system as a multi-agent platform: a Creative Production Agent for content, a Growth Agent for campaign optimization, and a Data Intelligence Agent built on what the company calls proprietary vertical data, "millions of car buyer personas, an exclusive knowledge graph, and tens of thousands of visual assets." The vendor's own framing leans hard on the integration of vertical data with Google Cloud's foundation models, in the language of "Agentic AI," rather than on a base large language model deployed raw.
What the announcement does not contain is the part that would let a buyer evaluate any of it. The release names zero specific automaker brands, dealer groups, regional markets, deployment counts, or signed customers. It speaks only in the abstract, in terms of "automotive dealers" and "globalizing automakers." Independent lookups for Aimotion on the Google Cloud automotive solutions page and across Google Cloud customer-story and blog archives returned no Aimotion-specific entry. LinkedIn and Crunchbase pages surfaced no traction either.
That is the structural shape of an evidence gap, and it is a pattern worth naming. In 2026, any AI marketing vendor can ship a press release. Most do. The market has learned to read for the second sentence of those releases, the one with the logo of a real customer, the geography, the baseline against which the headline number was measured, and the name of a third party willing to stand behind the result. Aimotion's release has none of those.
For an automaker considering any comparable pitch, the procurement-level questions are narrow. What was the 4-hour workflow before, and on what stack? What counts as a "high-quality" output for the 93% claim, and who audited the labeling? Where do the "millions of car buyer personas" come from, and can the dataset be inspected? What is the conversion rate, the cost per lead, the sales-accepted lead share, and the dealer-level close rate after a campaign runs through the platform? And finally: which named customer will go on the record, with which auditor, against which baseline? A Google Cloud partnership answers none of those questions. It tells a buyer only that the vendor runs on infrastructure a procurement team has heard of.
Aimotion CEO Hongyu Zhang, the only named spokesperson in the release, frames the company as a vertical-AI builder for high-ticket, long-decision-cycle industries, with automotive as the lead vertical. The pitch is plausible in shape: dealer marketing is exactly the kind of high-touch, multilingual, asset-heavy funnel where specialized tooling can win. Plausible shapes are not products, and partnerships are not references.
The thing to watch next is whether Aimotion produces a named customer, a third-party benchmark, or a Google Cloud customer story in the back half of 2026. If it does, the 30x claim becomes a real data point. If it does not, the claim stays a press release.