Your apps just got smarter. Did your policy?
We need to update our AI policy.
There are major changes that we really need to look at.
And if you're a smaller business and you don't really have a formal AI policy, that's OK.
You need to at least think about these topics, make some decisions and have a conversation with your team about them.
Let’s jump in.
Data boundaries
The first update we need to talk about is data boundaries.
All these AI tools are now getting into other apps, and other apps are opening their doors for the AIs to come and do its AI magic on their data.
Inside tools like ChatGPT, you can now run research across your Google Drive, Outlook calendar or email, SharePoint, Teams and many other platforms.
Should you allow that?
Well, in theory, it should be OK if you’re already allowing your team to upload that kind of information into ChatGPT or whatever tool you're using.
But in reality, you're now opening the doors much, much wider and allowing AI to look at everything.
Are you comfortable with that?
You need to make a decision.
You need to talk to your team about it.
AI agents with Browser Access
The second update relates to capabilities like the new agent mode in ChatGPT.
Agent mode means that the AI can now open its own browser and start clicking around the web. It can log into apps on your behalf using your credentials or limited credentials and go do things for you.
Are you happy for your staff to use this?
That requires a lot of thinking.
If it’s just searching the web to find restaurants or compare washing machines, the risk is very low.
But as soon as it starts logging in and doing things on your behalf, the risk spikes into the realm of unknown.
Right now, we shouldn't allow it to do anything logged in as us unless we're able to sit and watch every step it takes.
One thing you can do is give it access to apps through its own account, which has highly restricted access.
For example, if you're giving it access to your accounting system, create an account with read-only access so it can just pull reports.
We don’t know the risks yet. We don’t know what’s going to break. We don’t know where it’s going to get messy.
If you are going to let it do things on our behalf, only do it if you can watch it.
Otherwise, probably not.
Creating agents
Platforms are now making it very easy for us to bring in agents or create our own.
And I’m not just talking about ones like ChatGPT that literally click on screens.
I mean agents that have a saved prompt that gives them a personality, and sometimes even tools, like the ability to read and write emails, send messages or create files.
In Microsoft Copilot, for example, there’s an agent store.
Some agents are created by Microsoft, and others by third parties.
You can also create your own, assign it a personality, and add tools using Copilot Studio.
Other platforms like Zapier and Relevance also let you create these kinds of agents.
Are you going to allow that?
I’d say yes, for our learning, but only in an environment that’s well educated and well informed.
Microsoft’s agents from the Microsoft store? Not a problem.
Third-party ones? You'll need to take a long look at them to understand what data they’ll access and what they’re going to do.
If you’re building your own agents, I wouldn’t use them to interact with clients externally unless they’re very well tested.
If you’re just building them in-house to play around and learn, that’s fine. But you need to understand they’ll be a bit unpredictable, they’ll take time to become useful, and your users, even if only internal, need to know that.
Even if someone builds an agent that merely helps review NDAs, contracts, or onboard supppliers, everyone needs to understand that this is experimental, unpredictable and very, very new.
Time to review your AI policy
These are the significant changes that came up in the last few weeks.
And if you haven’t reviewed your AI policy since then, you need to update it.
If you don’t have an AI policy yet, or if you want to understand the basics before getting into all these updates, we’re going to run a session just on that.
Reply to this email. I’ll make sure you’re informed.
—-
Inbal Rodnay
Guiding Firms in Adopting AI and Automation
Keynote speaker | AI Workshops | Executive briefings | Consulting CIO
Want to receive these updates straight to your inbox? Click here: www.inbal.com.au/join
When you are ready, here is how Inbal can help:
Bring your entire team on the AI journey in just 30 days. This program is designed to give your team a solid foundation in using generative AI in responsible and impactful ways. Inbal helps you choose your AI tools, create an AI policy and train your team.
Want the confidence to set strategy and lead but don't have time to keep up with all the changes in tech?
Tailored for your needs, Inbal will works with you through one-on-one sessions to develop your technology literacy and keeps you up to date.
For CEOs, partners and business leaders. Everything you need to know about AI without the noise. Inbal shares the state of AI, recommends tools, and answers your questions about strategy, implementation and safe use.
Only what's real, no hype, no noise.
This is a one-off session for your entire leadership team.
