AI Firms Must Play Fair: Why Trust Matters With New Tech

This week, a big AI company called Anthropic had to fess up about something a bit sneaky. They were quietly putting hidden rules into their AI, called Claude, which stopped it from helping create competing AI models.

Think of it like a new oven that secretly stops you from baking certain recipes if they're too similar to a rival brand's. Researchers found out and called them out, rightly asking for more transparency – that's just being open and honest about how things work.

Anthropic has now backed down and changed its policy, which is a good step. But it highlights a much bigger issue for all of us, especially Aussie small business owners thinking about using AI. When these powerful AI tools have hidden limitations or biases, it can be hard to trust them.

This isn't about highly technical stuff; it's about fair play. If companies aren't up front about how their AI works, or what it can and can't do, it could unfairly impact competition or lead to unexpected problems down the track for businesses relying on these tools. It’s a good reminder that as AI becomes more common, we need to expect honesty and clear rules from the companies building it.