Australian Student Game Challenge to End as Generative AI Complicates Judging

The Australian STEM Videogame Challenge, a long-running competition that gave school students an opportunity to create and present their own games, is set to end after this year. The reason is not a lack of interest in game development. Instead, organizers say the rapid spread of generative AI has made it increasingly difficult to determine how much of an entry was actually created by the students submitting it.

For 12 years, the challenge invited young people ranging from elementary school through high school to make games around an assigned theme. Students could enter alone or form teams of up to four people. The completed project was not simply a game file: participants were also expected to provide a design document and a short trailer demonstrating their work.

That structure made the event more than a simple contest for a finished product. It gave students a reason to plan, discuss ideas, solve technical and creative problems, and communicate what they had made. But in an environment where text, artwork, code, audio, and other assets can increasingly be produced or heavily assisted by generative tools, the challenge's original approach to assessment has become much harder to sustain.

Why the competition is ending

ACER, the organization behind the challenge, has said the decision stems from the fast-moving and complicated arrival of generative AI in game development. Cathy Zemaitis, head of ACER's Student Learning and Progress division, said the technology was making it harder to judge what students had created themselves and was pulling the initiative away from its intended purpose.

For useful background on this topic, read Why Capcom Started Its Modern Resident Evil remakes With Resident Evil 2.

"The decision was driven by the rapid and complex rise of generative AI in game development, making it harder to judge what students made themselves, and moving the program away from its original purpose," Zemaitis said.

That issue gets to the core of what a student-focused development competition is meant to reward. A polished result is only one visible part of creating a game. The learning value often lies in the process: turning a rough concept into a workable plan, testing a mechanic that initially fails, learning why an asset does not function in-engine, dividing responsibilities among teammates, and revising a project after feedback.

Generative AI can be used in many different ways, and not all uses are identical. A student might use a tool for brainstorming, proofreading, temporary placeholder material, or technical explanation. At the other end of the spectrum, an entrant could submit work substantially produced by systems that generate code, visuals, dialogue, music, or promotional copy. Establishing a clear boundary between assistance and authorship is difficult in any setting. Doing so fairly across entries from children with varying resources, experience, and documentation is harder still.

For contest organizers, the problem is not merely detecting whether AI appeared somewhere in a project. A game is a collection of interconnected work: design decisions, programming, art, sound, writing, testing, and presentation. An entry may contain original work alongside external assets, tutorials, templates, and generated material. Deciding whether it meets a rule is one challenge; determining whether the final judging process remains equitable is another.

A loss for students who built projects together

The closure has been especially disappointing for families and returning participants who treated the competition as a meaningful annual creative goal. One parent, writing online about a teenage son who had entered several times with friends, described the practical and social value the event had offered.

"My teenage son and his friends entered the competition for several years, meeting up regularly, discussing ideas, and working together on a shared passion project," the parent wrote.

Those meetings represent an important part of youth game development that can be lost when discussion focuses only on the final game. A team project can teach collaboration in a setting where everyone has a tangible stake in the outcome. One student may be more interested in coding, another may prefer level design, another may want to draw or write, and another may become the person who keeps tasks organized. Even a small game can give each participant a visible role.

The parent also said the challenge gave the group confidence and helped them feel connected to a wider community. For young creators, that recognition can matter as much as an award. Finishing and submitting a project establishes that game development is not an abstract ambition reserved for distant studios. It is something a student can attempt, share, improve, and discuss with peers.

In a later reflection, the parent said the teenager was visibly upset by the news and felt that something tied to his identity had been taken away. That response underlines why the end of a competition can carry more weight than the loss of a single extracurricular activity. A recurring deadline and a shared event can provide structure for a hobby, especially for students who may not have a local club, formal course, or other outlet for their interests.

The difficult question of real-world skills

The challenge has described one of its purposes as helping students develop real-world skills. That goal now sits directly beside the debate over what skills educational programs should foster when generative tools are widely accessible.

Professional game development is likely to keep grappling with AI-related questions, from automation and asset creation to authorship, labor, quality control, and disclosure. Students will need to understand the tools that exist around them. But familiarity with a tool is not necessarily the same as developing the underlying creative or technical knowledge required to evaluate its output.

A generated code sample may appear useful until it breaks a game system or introduces a problem the student cannot diagnose. Generated art may be fast to obtain but still fail to fit a game's visual language, communicate a mechanic clearly, or work correctly within an interface. Generated writing may fill a dialogue box, yet not establish tone, character, pacing, or player motivation. Learning to recognize those shortcomings requires judgment, and judgment is built through practice.

This does not mean every educational setting must prohibit every AI-enabled tool. It does mean that an event built around assessing student work needs rules and evidence that allow judges to see the student's own contribution. If a competition cannot confidently do that, awards risk reflecting access to generation tools or the ability to present generated output rather than the development skills the program was created to encourage.

Could future student contests take a different approach?

The end of the Australian STEM Videogame Challenge does not settle the broader question of how youth game-development events should operate. It demonstrates how quickly existing formats can be strained when technology changes faster than rules, judging standards, and verification methods.

Future programs may choose to place greater emphasis on process rather than only on a final downloadable game and trailer. Development journals, version histories, team interviews, playable prototypes created during supervised sessions, and presentations in which entrants explain their design choices are all possible ways to make authorship more visible. Such methods would also give judges a fuller picture of learning, including experimentation and revision rather than just the sheen of a finished submission.

Some contests may establish explicit categories or disclosure requirements for AI use. Others may limit the tools permitted for a particular event, ask students to document generated material, or build challenges around live design tasks. Each path carries trade-offs. More verification can mean more work for students, teachers, volunteers, and judges. Tighter restrictions can be difficult to enforce. Broad permission can make comparisons between projects less meaningful.

There is also a risk in treating young people as if they are solely responsible for solving a problem created by an industry-wide rush toward generative systems. Students are navigating tools that adults, companies, schools, and platform holders are still debating. The disappearance of a valued competition shows that the consequences of that uncertainty can reach well beyond commercial game studios.

An ending that should prompt a wider response

For the students who used the challenge as a chance to learn and work with friends, its final year marks the loss of a rare public venue for their projects. They can still make games, share prototypes, join communities, and learn independently. Yet an organized competition provides motivation and recognition that are not easily replaced by simply telling young people to keep creating on their own.

The central lesson is not that technology can never be part of learning. It is that educational opportunities need structures that protect the value of effort, experimentation, and authorship. When a contest can no longer distinguish between a student's developing skills and work produced elsewhere, its ability to reward learning is weakened.

The Australian STEM Videogame Challenge was designed to help young people build games and gain confidence in the process. Its closure is a reminder that preserving spaces for beginner creators will require more than access to new tools. It will require formats that still make room for students to show what they can imagine, make, explain, and improve for themselves.