New AI Picks Trustworthy Online Collaborators — and Admits When It's Unsure
Could save you from bad freelancers and flaky business partners online.
Imagine you're hiring a freelancer or choosing a business partner online. You only have their past work to go on, but that work happened on different websites, devices, and projects. Some feedback is glowing, some is mixed, and some is just missing. How do you decide who to trust?
A new research paper from IEEE GLOBECOM tackles exactly this problem. The method, called multi-view evidential learning (MVE), treats every past interaction as a separate "view" of that person's reliability. It looks for patterns over time — like whether someone turns in work late only when the project is big — and combines all these clues into one clear score. What's new is that the AI also flags its own uncertainty. If it can't tell if someone is trustworthy because there isn't enough data, it says so instead of guessing.
The system was tested against older methods and consistently picked better collaborators, leading to higher task success rates. This matters because more and more work is done through online platforms, and automated systems are starting to choose who gets hired, who gets paid, and who gets promoted. If those systems can admit when they're unsure, humans can step in and make the final call.
The catch? This is early research, so it's not running in any real app yet. But the idea of AI that measures trust — and its own limits — could soon make online collaboration fairer and safer for everyone.
- The AI looks at a person's past work from many different sources, not just one review or rating, to judge if they're reliable.
- It uses a technique called evidential learning (being honest about uncertainty) — so it tells you when it doesn't know enough, reducing blind trust in AI verdicts.
- In tests, the system chose more dependable collaborators and completed tasks more often than existing methods.
- This could lead to smarter hiring, fairer online marketplaces, and safer automated teamwork.
Why It Matters
It means online collaboration and hiring could become safer, with AI that knows when to trust and when to ask a human.