AI 4 Sales: Guardrails and Ethical Considerations

Foreign. So let's cut to the chase. Without ethics there is no trust. And without trust in sales in particular, basically there's nothing. Now AI is exponentially complicated, both ethics and trust. And that's why we created that individual sales 300 labs to focus solely on these issues. And today we're fortunate to have a fantastic panel of experts that will be our first panel of hopefully many through our AI Sales summits. So joining us today are Julie Strauss, VP of AI Sales Marketing Incubations for Microsoft. Larissa Zuiter, who's with the board, is a board member of the center for AI and Digital Policy and our good friend Roderick Jefferson, who's enablement leader, keynote speaker and best selling author. Unfortunately Andrea Ruotolo had a last minute family emergency, cannot make it today, but her thoughts are with her. So I wanted to kick it off with really kind of setting a baseline of how we see the definition of ethics in AI. And I'll start with you Julie, if you want to take it away and also give us a little bit of.

Your background, I would love to. Thank you. Very exciting panel here. So I'm excited to the conversation we are going to be having. I'm a Vice president at Microsoft. My team is really focusing on developing what we traditionally thought of as CRM. So our sales and marketing applications now more or less renamed to AI Sales and marketing. And so it's really where we are investing for, for sellers and the sellers to kind of do their day to day work. And at Microsoft we really believe that when you create technologies that can change the world, we also must ensure that the products that we are releasing and the technology that we are using is being used responsibly. And for Microsoft that means a lot of things. And we have like a framework that actually guides every single product we develop when it comes to AI. So our core principles is really about fairness, reliability and safety, privacy and security, inclusiveness, transparency and accountability. And every single feature we develop go through that lens from every single team. And the goal is really to make sure that we maximize the benefits of AI and minimize any potential harm that can be done. So everything is going through a lot of scrutiny. But I think for me personally it's not just about how we build products, it's also what, what part of the sales cycle are we empowering or using AI. So if you think about sellers, right, what is unique for that, it is that human first quality. That is what a seller does. You have to have empathy, you have to understand strategy. You need to deeply connect with customers and understand their needs. We can never lose sight of that. So where do we apply AI? We apply AI where that has the greatest potential to improve that seller, to make that seller more successful. And that's really what we focus on. That will be stuff like making sure that AI works when you are asleep. Like make some of the research that can be done under the covers available 24 7. They're just some examples of really empowering the seller to be successful. So we really think of AI as assisting sellers at augmenting their work and make them kind of more successful in their work.

Right. Great. How about Larissa?

Yes. Hi everyone. So my name is Larissa Zutter. I'm from Switzerland. Currently I work as an AI growth and success manager at a Swiss AI startup called Peak Privacy. And I work on helping SMEs and NGOs implement generative AI by empowering their teams and their work streams in an ethical and responsible manner. I'm also a member of the Board of Directors at the center for AI and Digital Policy where I've been working for the last four or five years. I've also worked in various different roles in relation to ethical and responsible AI, both in consulting as well as in pharmaceuticals and fintech. My background is in international security and international political economy, so I have a social sciences background. But I have really enjoyed getting to delve into the more technical side through my work. I think for me, ethics in AI and this kind of pertains to any industry, not just or any function, not just sales. It really means considering what moral principles and values should be guiding both the development as well as the deployment and the use of of AI tools or systems. I think it's about really being clear within your organization about what you consider right and what you consider wrong. And also, just as Julie said, how to minimize harm whilst also maximizing the benefit that you can gain from these technologies. And I think also on a personal level, for me it's really about thinking critically about who gains from these new technologies and how these new gains are divided, but not just in terms of of profit, but also in terms of who benefits from the productivity increases or the upskilling that we make available to our teams internally. So making sure that the learning that you're able to gain also from using these new tools are spread somewhat evenly within the company so that you don't end up just improving certain jobs. So for me, on a personal level, it also entails thinking further than just the technology, but also what consequences it has within the organization. Organization.

Roderick hello everyone.

Roderick Jefferson. I'm the CEO of Roderick Jefferson and Associates. My background is in sales and sales enablement. For the last 30 years, I've been with a couple of small companies at&t eBay, Oracle, Salesforce, Marketo, etc. And I look at this as an amazing productivity tool. I look at it no different than a word, PowerPoint, Excel. It is a productivity tool that is only the outcome is only as good as what you put into it. Now, around the ethical piece, I think from the lens of a salesperson, this is extremely important. And I think we're at a tipping point right now, similar to where we were with GDP when it came to data. Right. There has to be guide rails, otherwise this thing can go a number of ways. Right. From the perspective. And you said something earlier, Jeff, that really grabbed me and that that word was transpar and trust in sales. Right. And let's be honest, not everything about sales right now has a positive slant. And I think there's an opportunity to utilize this tool to one, gain some credibility back and secondly, to really help drive productivity and profitability in a way that we have never been able to do in our previous lives. Because the access to data and information is so prevalent. But at the same time, to double tap on what Larissa said, it also comes down to who has access to the data, what happens with that data, and how it's being utilized and leveraged. And so it's important that we look at it from all angles from the perspective of you, the user, the individual, also the folks that are consuming that content. We look at it from a governmental perspective. We also look at it from a company expectation as well as ownership and accountability perspective as well.

Right. And Julie mentioned something I thought was very interesting because when she described Microsoft's philosophy, it sounded like it start. It didn't start with creating a tool and then figuring out how to inject ethical guide guardrails into it. It sounded like, and correct me if I'm wrong, but it sounded like you have a filter where everything goes through that kind of ethical process first before it goes into the, you know, the steps that go into production and then filtering out to your clients and your customers. So I'm wondering, and maybe, Roderick, you can start with this one. Should we like for our AI Sales Summit here, we have over a thousand sales leaders that are looking for AI solutions that are going to improve their productivity, improve their companies, their revenue growth. Is it enough that we have companies like Microsoft that are, you know, trying to do their best at creating ethical solutions? Or, or is it the responsibility of us as sales leaders when we're selecting products? Should we be looking at that? Should we be starting from the place of looking for ethical solutions as well?

I'd say the answer is absolutely yes, unequivocally, because, and I'll start with a very simple example. We're all on social media, right? And we know that, especially in the US Coming off of a presidential change last year, regardless of where you sit on the aisle, I could care less about that piece. Right. But more about the fact that bias has become far more important and infused, I think, than I've ever seen before. And from an ethical oversight perspective, to answer your question directly, I think that it's not just the companies that have that responsibility. I think that the companies, as I said earlier, the government has a piece now, at least from a consistency perspective, not from a determination of how you use things. And I also think we as users need to be very, very focused on where your data, your information is going and also the kind of information that you put out there as well. Because once it's out there, we know there is no deleting. It's going to be out there and it's going to take on a life of its own. And I think from my experience, especially on the governmental piece, I think that relying solely on the government to regulate leaves companies kind of scrambling to catch up. And the evolution of AI is happening so fast. I don't know that that's where it should lie because I don't believe we are the speed of business. We're not the speed of light anymore. We speak now at the speed of AI and that thing is changing. So what can be put out now for guidelines today, a week later may not fit. Right. I think the most effective model kind of combines the proactive self governing industry, more of wide standards that are being developed through things like, you know, trade associations and regulatory frameworks. But I think that each sales organization should have an internal, excuse me, internal AI ethics committee that are participating in outside the walls of where they are and a broader industry initiative that then establishes a set of shared best practices.

And that's what we're trying to work on in sales 3o labs as well.

Absolutely.

Larissa, did you want to add something there?

I mean, there's really not much to add from my side because I fully agree with what Rodrick just said. I think it's so important, and this doesn't just apply, I think, to AI, but it's so important that you as a company don't outsource your ethics like to either government or to a company. It's so important to have values that you as a company believe in, especially now where there's a lot of divide in the world, not just in the US And I think it's super important to have internal, internal kind of understanding of also what external values you agree with and what you don't agree with and what you want to be part of and what you don't want to be part of. Because there are so many different frameworks nowadays. There are so many different trade organizations making ethical guidelines that you also as a your own organization need to kind of like Rodrick said, have your own ethics committee or whatever you want to call it, but that, that is able to pick and choose. Okay, what do we want to agree with? What don't we agree with from this, this other organization? And think it's incredibly important to be proactive about it, just like Rodri said, because governments are so behind in regulating this. And I think especially in an area like sales, where trust is so absolutely essential for your business, that it's super important to have a certain degree of integrity and kind of know where you stand to some degree in order for your customers to also be able to. Yeah. Give you that trust. So I think from that standpoint as well, it's absolutely essential that you as an organization have your own compass in that sense as well.

Right. I love what you said. You can't outsource ethics. And, and Julie, how does Microsoft, you know, look at the. That kind of parameters? As far as I know that, you know, you have the government regulations and potentials and things like that. And I. But I know that like your CEO of AI opines quite often on LinkedIn about AI and ethics and kind of existential type issues.

Yeah, no, I think agree with everything that has been said to be on. That's also why we have the frameworks that we have. And I think for Microsoft, I think we've been strong advocates of having kind of law and governance over it, but also really encourage companies to have their own ethics. Like Larissa, you said, I actually really love that. And ask those type of questions from your vendor, like have the transparency. In some examples, if you send out kind of communication to customers that has been AI generated, make sure you have full transparency of what is being communicated, what are the decisions or what are the suggestions that AI is making. And I think for me personally, the latest incubation I've been working on, you ask a question, the agents go to Work and bring an answer back at any point in time. You can kind of go to the backside, you can see work and you can see all the reasoning for why, why did we make the suggestions. You can even go to what's the SQL query that was written to get this number. And that is a very high bar, but that's what we need to get to. So before you really dig in and deploy externally that you feel super comfortable with what is being done for you through AI. So I fully agree with everything that has been said. Microsoft, from legislative perspective, like super committed to working with officials as well to make sure that we help also share our learnings because we learn also from ethics, from customers every single day in, in what they expect. And we are as a company fully committed to bringing that back to the community as well. So.

And that's so important because I mean, like, you know, you have some people, every once in a while you'll hear, well, maybe more than every once in a while. AI is like magic, but it's not. It's a computing tool and to be able to know exactly how it works and be transparent about exactly how it works benefits everybody. And we have an interesting question from the audience, which is who bears responsibility when AI driven sales recommendations lead to customer harm or discrimination from either the salesperson, the manager or the technology provider itself. Who wants to take that? I'll throw it out there for anybody who wants to jump in on that one.

I'll jump in on that one. I think it's a shared responsibility and that is why it is so important to make sure one, the information that's being shared is done transparently and that it's also pressure tested, if you will, to make sure that it's accurate as well. Because we know that there are random things that can show up. And don't just assume because it's shared with you that it's accurate. We all know that, right? If you've ever, you know, put in a prompt about yourself and realize that, oh no, I didn't even go to that school, or that's not what I did or how I did it, then that happens over and over. And I think to your point, or to the question of why it's important for shared, because I don't think we have enough experience yet to say that we can point a finger. I think we're in those, one of those situations where if you're pointing a finger, you know, three more are pointing back at you. So there has to be a way for all of this to to be owned and shared. And I look at it kind of in three pieces and that's communication, collaboration, then orchestration. And in the communication being transparent and being clear. Because I've, you know, seen sales teams that got burned by implementing AI tools without clearly communicating to the prospect one, that AI is being used. And second, your customers deserve to know honestly and they'll respect you a lot more for it. Right? And it's all about how it's happened. Now the orchestration piece is all about all three of them working. When I say them to the question, the salesperson, the manager and the technology provider to make sure that you're in sync and that you are communicating the same way consistently out. And then finally the orchestration piece and I think that goes above and higher than any of those three pieces. It starts at the top of the company and it starts with the CEO, the cxo, the the E staff, right? Because they are the ones especially now with, with AI fluency and also AI implementation. They own culture of a given company and that company now will be seen as either honest and transparent or not trustworthy. And to me they what I mean by setting that culture, culture is what happens when no one is watching. And it's not what you put in your marketing, it's not what you put out, it's what your company does. Right? And not just what's happening in today's news cycle, but those personal experiences, whether they be positive or negative, again falls back on leadership.

Julie, did you want to add anything earlier?

I agree. I think it is a shared responsibility. Like for us as a vendor, we are responsible for making sure that we build the products responsibly, but also that we share with customers how we do that and what the ethical use of the technology is. They can't just be on one. It is really a shared responsibility. And back to Larissa's point before really making sure that you have your own kind of expectations and guardrails for what you want to send out. So it is really that joint responsibility across the user, the manager, the company and the software vendor. And obviously regulations on top of that. But to earlier point, we can't wait for that. Make sure we all take accountability in that end to end cycle.

And there was a follow up what level of AI transparency. And part of this I think was sort of answered but what level of AI transparency should salespeople provide the prospect should buyers know when they're interacting with AI enhanced communications? And we see that a lot. I mean the, you know, auto generated emails that Go and scan your LinkedIn and then say, hey, I see you're doing this. When the person themselves that's sending it doesn't know they're doing that and they didn't bother with it, but they're having an auto tool to do it. Larissa, you want to speak anything to that or.

Yeah, sure. So I mean, this is obviously a personal choice. I personally fully think that you should. So I fully agree with what Roderick said before about if, especially if you're working in an industry where trust is so important, why would you like risk that in the first communication by not being honest about that? I think you almost gain more respect by being honest than by trying to like cover it up in that or not cover it up, but not be forthright about it. Right. So I personally think you can't gain much from or you can, you can lose so much from not disclosing it and then it being found out or the person realizing. So I think just being honest about these things and disclosing them. And this doesn't just go for like disclosing external communications with, with, with prospects, but also just in general, like when we work with SMEs to implement gen AI solutions, we're very honest about the limitations. Like there are limitations of AI. Like there are. And also within your company being very, very investing a lot of time in really making employees understand there are limitations. You have to. Look, this is not like a solution where you just, you, you completely get to take your hands off and you let it run free. No, there's like shared responsibility around everything. Right. So you have to be a very informed user and, and know the limitations of what you're using.

Yeah. You can't over bill it as a time saver, but by cutting out any human interaction and letting the auto AI go for it.

Exactly.

Jeff, I'd love to add something to that.

Yes.

Just come really grab my attention on the, the last piece. Human oversight isn't optional. It's strategic now. And it's a must when it comes to AI. Right. Every AI generated email proposal recommendation must, must, must be re reviewed by a human. Not just for clarity, but for credibility also because to your point, something you just said, you can send something out and go, oh, that's not really how we do it or what we do it. And now think about what you've done, all the work that you've done as a sales professional, you just completely killed all credibility. And so by simply taking some time and checking for not just content, but context as well as Honesty and clarity and transparency means a whole lot more now, I think, than it ever did.

Right. And that's. That's actually, you know, it's not only the right thing to do, but it's also good marketing to be able to say that, you know, we're doing things the right way. You're already seen. I mean, there's a, you know, the Ted Danson phone commercials. There's already one where it's like, oh, we. You talk to real people, not AI agents when you call our call center. So you're already seeing kind of a counter marketing. You know, to every Beatles, there's a Stones. So to AI, then there's human, you know, kind of pushback.

I saw something interesting just yesterday in a response I got back from someone that I had never seen. They had their footer with all of their usual content, and then on the bottom, the line said this email response was 100% written by a human. I have never seen that before. Never even thought about having a need for that until now.

Yeah, actually was thinking about that because I was thinking in the terms of. I've. I've seen it in some of the ethics associations, and it reminded me of, like, the Good Housekeeping seal of Approval. So, I mean, it has a place. But there was one other question coming through. Are there any industry coalitions around finding and sharing where there may be misinformation that should be eliminated from AI consideration? I don't know. Maybe Larissa would have something on that. I don't know. I don't know specifically.

So to be honest with you, I'm not aware of, like, I don't really know the sales industry coalition landscape. And I also don't quite understand the question because I don't know what's meant by that should be eliminated from AI consideration like that you're eliminating it from your choice of vendors or. Right.

Yeah, I'm not sure.

I don't quite understand the question. So maybe whoever asked the question could please specify so we can actually get you a good answer.

Right, right. And one of the last things I wanted to cover because we're running short on time, but I. I want to ask if you think that AI ethics is kind of. That AI is accelerated so quickly that we're kind of behind in AI ethics. We have catching up to do, and if that's the case, what are some ideas on how we can get caught up to where the technology is? And Julie, maybe we'll start with you.

Yeah, I go to Microsoft, I can start. And I think Speaking on behalf of Microsoft, I. It's not a sales pitch, but I frankly feel we were, we tackled this upfront, like we did the same when moving to the cloud. And I feel all these frameworks we put in place before we were allowed to release anything. The first cycles of AI we shipped took a long time because we have this full board of assessment and evaluations we had to go through that took weeks and weeks and weeks. So I feel honestly we've put in a lot of work. Way ahead of the game. This, like in this wave of technology transformation that I feel in a good state. But. But again, as a consumer of AI, I do definitely think there are areas where we, some products may have gotten ahead of their skis, so to speak. But I think, I think there's a lot to be learned and lot of kind of learnings that many vendors can tap into. So it's a little bit of a mixed response from that perspective, but I would say I feel pretty good about the kind of the offerings that at least I work on day to day. We have a lot of guardrails around, around that.

Right. And how about Larissa?

Yeah. So I don't think that AI ethics is behind, but I think that the operationalization of the principles are definitely very behind. Like, I think I've been working in this space for I think four or five years now. So before ChatGPT was released and this, the principles have stayed the same. Like, no one's reinventing the wheel right now. Like, everyone has very similar principles. There's a lot of convergence around. Like there was an article, I think, in 2021, around frameworks and what principles they were recommending. And it's this, they were all converging around the same ones that are still being used. So I don't think that we don't have a set of moral principles that we think are good. I think what we're super behind on and will probably always be behind on, unless we integrate ethics into the very beginning of development is actually how do we operationalize it? Like, of course the transparency is amazing and nice and everyone wants it, but how do you. How does that work? Like, how are you actually going to make that part of your product development? Or like, how do you actually infuse all these ethics into a product that you're making? So I think that part is very much behind. I don't think that the ethical principles that we've developed are per se outdated. I don't think you need to reinvent the wheel for every new technology I Think it's about figuring out how to apply it to every single technology.

That sounds like it goes back to what Julie was saying, what Microsoft has been doing for a long time. Roderick, last words.

Yeah, I agree with both of the ladies earlier. I also want to kind of add in. I feel like what I'm seeing right now is we still, because the technology is evolving so fast, we still have a moral and an ethical kind of slide rule, if you will, that shifts from day to day. And until we get our arms around that piece, that's where we are. And I think the problem is again, the speed of AI. As soon as we say okay, here's a standard, then new technology is added on top of it or new products are added on, then you kind of go, okay, well we didn't really think about that piece. So now we've got to talk about that next phase. And I think that's something that's going to be an ongoing piece because as fast and as much technology is happening, we're going to continually need to edit. But to what Larissa said, I don't think we need to completely overhaul and throw the baby out with the bathwater if you will. But what we need to do is make sure that as those new technologies are being released that we're prepared for those and we're open minded on how we're going to approach those.

Right.

Can I just add one thing that I love that I think one thing that is really important that I think is changing as well. We can't just as a vendor look inside our own organization. It is learning from the community as well, as well as sharing back so we don't all have to learn the same lessons from the beginning. And I think we actually spending a lot of resources giving back and learning from the community. And I think that's one, one many steps to really make sure we, we get ahead of, of a lot of those learnings as well.

And I think it's consumers. We have a responsibility to hold these companies ethically and morally right and we are taking and consuming this and utilizing those tools. We should have a voice and a vote in this, of the direction these things are going to. Otherwise we're going to get to a place where it gets really, really scary and out of control.

Right. And the boat is what solutions you're using. And that's one of the things we're working on. I want to personally thank every single one of you for participating and hopefully we'll have you back again because I hopefully this is a ongoing and expanding discussion that we're going to be having at the AI Sales summits. And to all of the participants out there, I encourage you to reach out to me if you want to discuss what we're doing with the AI research and AI ethics. And I thank you very much. And I'll turn it back over to Gerhard and Perry. Jeff, can you tell us about the Ethical Quality Index research that you have done? Yeah, we've created an EQI basically, where we utilize. Well, we utilize probably over 100 data points within five broad categories where we look at different aspects of the AI solution. So we started with chatbots and then we ranked seven major chatbots, put it through all of these tests, and then created an Ethical Quality Index on a scale of 1 to 100, with 80 and above being, you know, recommended or acceptable for use, between 50 and 80 needs work and then below 50 not necessarily recommended for use. So we've done that now for chatbots and we also did it for the legacy or the non AI centric CRM systems. And we're also now working on the AI first native AI CRM systems. And if you want to reach out to me, we also, I can also get you access to all of those as well. And you've also analyzed the revenue engines. Yeah, we did that on a, on a UQI or user Quality Index, and we're doing more work on that as well. Well, I want to thank everybody on the panel. Perry, last comment. What did you think of the overall to, you know, summarize it in one sentence, what you take away from this conversation. Trust.

I think that that's the key word.

I heard over and over and over.

Again, which I think is such a key factor in the world of sales.

But, you know, so much even more important as we enter and use more AI. So, yeah, trust, there's no sale.

So to speak. But I think, I think there's a lot to be learned and lot of kind of learnings that many vendors can tap into. So it's a little bit of a mixed response from that perspective, but I would say I, I feel, I feel pretty good about the kind of the offerings that at least I work on day to day. We have a lot of guardrails around, around that.

Right. And how about Larissa?

Yeah, so I don't think that AI ethics is behind, but I think that the operationalization of the principles are definitely very behind.