I’ve been curious given the outcome of the AirCanada ruling. I searched through the topic on HN (https://news.ycombinator.com/item?id=39378235) and some talk about the legal test of apparent authority. That was partly helpful in how that scope is likely to be tested in the court of law.
What are the actual mechanics within the company? For example, the companies I’ve worked for typically give delegation of authority to human roles that dictated what that role can or cannot do on behalf of the company. Some authority is even written directly into company bylaws, like the ability to enter into a contract with financial implications, or issue stock, etc.
Side note, on the AirCanada ruling. I find it hard to believe that company would go to court over a few hundred to a thousand dollars. Most customer service departments have the authority to just solve that issue without even tapping the legal team. My best guess is that they chose a case with minimal damages to test how the courts would react when an AI agent conflicts with the official policy of the company.
0 comments
—
Ask HN: How are AI agents given scope of authority at companies? · HackerTrans
What are the actual mechanics within the company? For example, the companies I’ve worked for typically give delegation of authority to human roles that dictated what that role can or cannot do on behalf of the company. Some authority is even written directly into company bylaws, like the ability to enter into a contract with financial implications, or issue stock, etc.
Side note, on the AirCanada ruling. I find it hard to believe that company would go to court over a few hundred to a thousand dollars. Most customer service departments have the authority to just solve that issue without even tapping the legal team. My best guess is that they chose a case with minimal damages to test how the courts would react when an AI agent conflicts with the official policy of the company.