Is “good AI” being defined by awards?

There seem to be a lot of AI awards appearing lately. Some focus on business impact, others on products and tools, and some are even recognising AI influencers and virtual personalities.

It’s easy to see why. New technology, growing ecosystem, a natural moment to highlight what’s working. But the more you look at it, the more it feels like something else is happening. Awards don’t just recognise progress, they start to shape what progress looks like.

What counts as “good AI”?

Right now, there isn’t a single shared definition. Some awards focus on measurable outcomes like efficiency, revenue or cost savings. Others lean towards usability, highlighting tools people actually use and enjoy. Then there’s a growing focus on creativity and attention, where AI-generated personalities and content are being recognised.

All of these are valid. But they’re also very different answers to the same question: what is AI actually for?

Awards as signals

That’s where it gets interesting. Awards aren’t neutral. They act as signals. They influence what companies build, what investors back, and what the rest of us pay attention to. If the biggest stages reward efficiency, we optimise for efficiency. If they reward visibility, we optimise for attention. Over time, those signals don’t just reflect the industry, they start to steer it.

Innovation or packaging?

There’s also a quieter tension underneath it. Some awards are grounded in real outcomes, with clear use cases and visible impact. Others sit closer to positioning, where visibility and credibility play a bigger role. It raises a slightly uncomfortable question: are we rewarding innovation, or the ability to present it well?

Still being defined

None of this is necessarily a problem. Most new technologies go through a phase like this, where progress, storytelling and noise sit alongside each other. But it does mean we’re still working out what matters, not just what AI can do, but what it should be doing.

And maybe that’s the more interesting shift. We’re not just building AI anymore, we’re shaping the criteria for what “good AI” looks like. And right now, that definition isn’t settled.