Trang chủInternational FootballAI in Vietnamese Football: The Money Flows to What Can Be Seen

AI in Vietnamese Football: The Money Flows to What Can Be Seen

**Core answer** Trí tuệ nhân tạo trong bóng đá Việt Nam hiện tập trung ở ba lớp: theo dõi khối lượng vận động bằng GPS, gắn nhãn video bán tự động và phân tích dữ liệu khán giả. Rủi ro lớn nhất nằm ở quyền hình ảnh cầu thủ trẻ và nghĩa vụ công bố khi câu lạc bộ dùng mô hình. **Key facts** - Phần mềm tự gắn nhãn pha bóng tại một câu lạc bộ V-League đạt khoảng 7/10 tình huống, phần còn lại xử lý thủ công. - Điều 32 Bộ luật Dân sự 2015 quy định quyền của cá nhân đối với hình ảnh, gồm cả việc sử dụng hình ảnh. - Nghị định 13/2023/NĐ-CP về bảo vệ dữ liệu cá nhân có

On the morning of a pre-season press conference, a V-League club's technical director told reporters that within three years artificial intelligence would change the way Vietnamese football recruits. He spoke without notes. That same week I sat in the club's video room. On the desk were 24 new GPS vests, a renewed contract for the performance analyst, and an A4 sheet of manual event-coding keys, yellowed from three seasons of reuse. The analyst told me the auto-tagging software got roughly seven of every ten events right, so he still clicks the rest by hand.

Seven out of ten. That was the entire AI footprint of that club on the day I visited, and it appeared in none of the slides at the press conference. The gap between the sentence in the meeting room and the invoice in the video room is what I want to measure here, because it determines what Vietnamese football can actually buy from the wave now breaking over it.

Globally, AI in football has moved past the exhibition stage. Major leagues use models to price tickets by demand, forecast injuries from GPS load, semi-automate video tagging, and generate content for social channels. Vietnamese football sits at a different point on the curve: thin data infrastructure, club budgets that differ by multiples, and a number of full-time analysts in the V-League you can count on one hand.

AI in Vietnamese Football: The Money Flows to What Can Be Seen

Results pressure runs the other way. The ASEAN Championship 2026 title, won on Thai soil in January 2026, pushed expectations up. VAR arrived in the V-League in the 2026 season and has expanded since, but VAR is broadcast technology. It shows viewers a line; it does not tell a coach who should partner whom.

Three application layers survive in Vietnam right now, measured by cost per unit of value: GPS load monitoring, semi-automated video tagging, and audience data for ticket pricing. None of them makes television. Nobody films a morning of data entry, and nobody interviews the person fixing mislabelled events. The real value of AI in Vietnamese football sits in the layer nobody films: clean data, clear contracts, and one named person accountable for checking the model. It is also the only layer a First Division club on a modest budget can still enter.

An international scouting-database subscription costs a few thousand dollars a year per account, a line item a second-tier club has to think about. Many teams send clips to a contact instead. AI does not fix that habit; it makes the habit faster and harder to audit.

Seven out of ten is the number I wrote down, and it means something concrete. If a team trusts the software completely, three of the ten most important events get the wrong label, and nobody knows which direction the error runs until the goal goes in. Translated for the television viewer: the next day's technical meeting will discuss a different problem from the one that happened on the pitch.

AI in Vietnamese Football: The Money Flows to What Can Be Seen

I have made exactly that mistake. In December 2026, analysing Morocco's 4-1-4-1 defensive block at the World Cup, I wrote that they won 14 of 14 aerial duels; the correct figure was 13 of 14. An account that audits data found it within hours. I deleted the post, rewatched the footage, and republished a correction within two hours with the error stated in the first line. Since then I cross-check every number against at least two sources and date the data inside the article. A model needs the same discipline, with one difference: it feels no shame.

Put the two side by side: a recruitment model trained on V-League data, where each player has few matches per season, few tracked metrics, and labels applied by the club itself. The model will answer with great confidence. That confidence is the risk, because it looks exactly like a hastily written scouting report, except with charts and better typography.

I still draw the diagram by hand before I open the software. The hand-drawn shape from the 2026 World Cup still reads tonight's match; what changed is speed, not grammar. The crowd watches the star; I watch the space behind him. In 2026 I wrote my first piece on a V-League match in which SHB Da Nang lost 0-3 at home to Ha Noi; all three goals opened down the left flank. The comment I received asked what a girl knows about football. I did not reply, I just added a map of each player's average position. Data was a shield then. Now data is merchandise, and merchandise can be counterfeited.

The unmarked pitch is image rights. Vietnamese football is used to selling collective image rights to sponsors and to game products. Once generative models are good enough to put a player in a club shirt in an advertisement he never filmed, the value of the original collapses through supply: unlimited replicas push the price per unit toward the cost of generation, close to zero. For a V-League player, image income is already thin. This is the first line to erode, before wages and before bonuses.

The legal basis exists: Article 32 of the 2026 Civil Code on an individual's right to their image, and Decree 13/2026/ND-CP on personal data protection, in force since 1 July 2026. Both predate the spread of generative models, and neither has been translated by any club I have asked into a specific clause in a young player's contract.

A 17-year-old signs a training contract with one line about image rights, usually without being told that the line covers using his face to train a model. A small detail. In three to five years it becomes a large case, and the person who pays is the person who signed at 17.

The most useful operational lesson from the AI disputes in the creative industries is that the penalty falls on the person who concealed, not the person who used. Director Eli Roth lost credibility because he had previously said the opposite of what he did, not because he used AI in a handful of scenes. The principle transfers directly. A club that uses a model in medical or scouting work without disclosure will lose more than it gained when the use surfaces. And an unstoppable trend does not stand up in court, nor does it substitute for a plan.

The counterintuitive part: the money in Vietnamese football's AI story is flowing to the visible layer. Banners, launch clips, signing ceremonies with a backdrop. The invisible layer, the people labelling data, the contract clauses, the cross-checking routine, is starved. I have seen this structure once before, in youth development. Over the past decade many former internationals opened academies, with fine opening ceremonies and photographs with children, while the number of properly trained grassroots coaches grew far more slowly than the number of signboards. An academy without a curriculum is still a rented pitch. A data centre without an auditor is still licensed software.

In Vietnamese esports I see the same failure shape. A women's tournament built inside a closed ecosystem, importing teams for the visuals, produced after three seasons not one player capable of competing in an open event. A beautiful stage, and nobody walking out of it.

Next season the test for each club is a single question: what does the image-rights clause for a player under 18 say about using his face to train a model, and who signs on his behalf? The answer tells you who is laying a foundation and who is hanging a banner. Tactics are not magic. It is just that some people look a little longer. Data does not lie, but it is good at hiding surprises. I will go and ask each club, date the answers, and publish them when they arrive.

Cầu thủ liên quan