What ethical responsibilities do software engineering leaders have when it comes to generative AI?

Make sure software is not biased and that the AI is well tested and accurate

Ensuring responsible AI that is as free from bias as can be, and that it is used for ethical purposes.

Ethical and cultural considerations. 
Avoid bias and discrimination in any form and cleanse the data

A good overview of ethical considerations for AI applications can be found in two ISO publications.ISO/IEC TR 24368:2022 lists a dozen or concerns, and it is fair to say that what is true for AI in general also applies to Generative AI. So refer to https://www.iso.org/standard/78507.html for more information.

But it is fair to say that generative AI has put the spotlight on ethical responsibilities such as accountability, fairness and non-discrimination, transparency and explainability, privacy, respect for the rule of law, environmental sustainability and labour practices.

Now that AI is capable of removing existing tech jobs as well as creating new ones, but quite possibly make a greater number of traditional roles redundant than creating new ones, AI tech leaders are confronted with having to make decisions that were previously 'reserved' for their business counterparts.

So if you want or need to understand which ethical responsibilities are carried by your organisation (and by extension by you): you could do worse than finding like-minded individuals who have dealt with ethical responsibilities in other parts of the business. In my experience these responsibilities are driven more by the industry you operate in than by the technology you deploy.

These developments are recent, the ethical considerations in your industry less so.

One issue is to avoid bias and make active efforts to recognize and minimize biases in AI training data and algorithms.