What's Your Manager's Expectations on AI?

Hello,

This is Simon with the latest edition of The Weekly. In these updates, I share key AI related stories from this week's news, list upcoming events, and share any longer form articles posted on the website.

For many of us, our use of AI at work is entirely defined by corporate policies and our manager's attitudes. In over a year of writing this newsletter, I haven't really touched on how much influence our managers have over how we use AI — and for which tasks. At one extreme, some managers have become near-obsessed with the technology, even tracking how much their staff use AI as a measure of efficiency. This has given rise to the eyebrow-raising term "tokenmaxxing." At the other end of the scale are managers who are paralysed by fear and heavily restrict their teams from using AI at all. Clearly, there's a healthy balance somewhere in the middle — and it's worth having an open conversation about it with your team.

In my own team, we use Claude on an Enterprise plan and make extensive use of it. My manager is always sharing new ways he's using it and encourages us to get the most out of it. Across my global team, people are constantly sharing new Skills they've built and new projects they're working on. It's a genuinely collaborative approach to using AI in our day-to-day work.

And if your team isn't really using AI yet but you are, this is an opportunity to show some leadership — share what you're doing, offer guidance to your colleagues, and demonstrate what's possible.

Have you had a clear conversation with your manager about their expectations around your AI use? If not, here are a few questions worth raising:

When you type a prompt, you summarize. When you speak one, you explain. Wispr Flow captures your full reasoning — constraints, edge cases, examples, tone — and turns it into clean, structured text you paste into ChatGPT, Claude, or any AI tool. The difference shows up immediately. More context in, fewer follow-ups out.

89% of messages sent with zero edits. Used by teams at OpenAI, Vercel, and Clay. Try Wispr Flow free — works on Mac, Windows, and iPhone.

Start flowing free

In this section, I’m going to bring to you a real world example of AI use. This week, we take a look at an infamous case of how Klarna implemented AI, but also made a u-turn.

A global survey of 6,000 employees at enterprise organisations found that between one-fifth and one-third of workers use AI outside the influence and governance of their IT function. The research documents a workforce split into two groups: employees equipped with IT-managed tools, training, and oversight, and those operating independently with consumer AI services. Many employees report that their employers fail to supply either AI tools or training, and a sizable share of those who receive training describe it as irregular or ineffective.

Why it matters: The gap between how fast employees are adopting AI and how fast organisations are governing it is getting wider, not narrower. Shadow AI was a factor in one in five data breaches last year, adding an average of $670,000 to the cost of each incident. If your organisation doesn't have a clear AI policy and proper tooling in place, employees aren't waiting around for one.

A landmark case making its way through the US courts could redefine employer liability for AI tools. The Workday AI lawsuit — now a class action — alleges that Workday's AI recruiting software discriminates against older job applicants and Black applicants. In early March 2026, a federal judge allowed the age discrimination claims to move forward, with a final decision expected later this year. The court's treatment of the software vendor as an "agent" of the employer is groundbreaking — it means that if a third-party tool makes a discriminatory decision, the employer cannot simply point the finger at the vendor. Both parties are now in the crosshairs.

Why it matters: Most organisations using AI in their hiring process assume the vendor carries the liability. This case suggests otherwise. If the tool screens out candidates unfairly, the employer is exposed — even if the tool came off the shelf and was never customised. "We don't know how it works, but the vendor says it's fair" is not, according to the EEOC, a legal defence.

The two biggest providers of workplace software decided, within days of each other, to give IT teams a proper view of what AI is actually doing inside their organisations. Microsoft's Agent 365, made generally available for commercial customers on 1 May, is designed to help organisations discover, govern, and secure AI agents operating across Microsoft, third-party SaaS, cloud, and local environments. Google's new AI control centre for Workspace, announced the same week, gives administrators a centralised view of AI usage, security settings, data protection controls, and privacy safeguards.

Why it matters: The fact that both companies launched governance tools in the same week tells you something about where the pressure is coming from. Organisations are no longer just testing AI chatbots — they're deploying agents that can reach into business systems and take actions on behalf of employees. That's a fundamentally different risk profile, and IT and compliance teams are starting to catch up.

Thanks for reading, and see you next Thursday.

Simon,

Was this email shared with you? If so subscribe here to get your own edition every Thursday.

Enjoying Plain AI? Share it and get a free gift!

If you find this newsletter useful, why not share it with a couple of friends or colleagues who would also benefit? As a special thank you, when two people subscribe using your unique referral link, I’ll send you my "Exclusive Guide: Supercharge Your Work with NotebookLM." It’s a practical, no-nonsense guide to help you turn information overload into your secret weapon.

You can also follow me on social media: