Workplace AI Is Moving Beyond Automation
Reading Time: 10 Minutes
Image created by Superumxn team.
Time
Most of the conversation about AI at work is still about saving time.
And fair enough. Automated admin tasks and meeting transcriptions are genuinely useful. If you’ve spent any time with Microsoft Copilot or Notion AI, you have probably already cut a few hours of grunt work out of your week. But there is something more interesting starting to happen alongside AI automation.
Some LLM tools are now watching how you work and prompting, or ‘nudging,’ you to do it better. Take a manager who has not recognised a team member's contribution in two weeks, who gets a discreet prompt to remind them. Or an employee who has been in back-to-back meetings all week gets a reminder to block some focus time. If you keep logging on at 11pm you might receive a gentle flag before it becomes a habit.
None of these feel like big changes to how we work. That is the point. Deloitte predicts nudgetech will be one of the top work trends in 2025 and beyond, and research from MIT and Stanford's Human Dynamics Lab backs this up. Why? Peer influence and well-timed prompts change behaviour 3 to 5 times more effectively than formal training alone.
Here are 12 companies already making it work, and what you can learn from each.
What can I expect from a subscription?
Real-world strategies for navigating the future of work.
AI AGENTS AT WORK
Everything you need to know about how AI agents are changing teams and ways of working.
SUPERHUMXN IRL
Real insights from investors and operators.
MEMOS ON PEOPLE, CULTURE & AI AT WORK
The latest use cases and strategies for change and transformation.
1. Google - Behavioural design before its time
Google was doing this before anyone called it nudgetech.
Project Oxygen started in 2008 with a fairly simple question: what actually makes a good manager? The answers were not revolutionary. For example, care about your team, coach more or don't micromanage seem like obvious suggestions. But Google did not just publish a report, they embedded the findings into tools managers were already using at work like checklists or onboarding processes.
Project Aristotle went further. Google analysed 180 internal teams expecting to find that the best ones were simply filled with the most talented people. They were not. The teams that consistently outperformed were the ones where people felt safe enough to speak up, disagree, or admit they had got something wrong, without worrying it would be held against them. Google called this psychological safety, and they also adopted these principles as cultural practices.
This is the foundation of everything nudgetech builds on. Create the right environment and performance follows.
2. Microsoft - Productivity tools influence habits
Microsoft is one of the clearest examples of behavioural AI becoming embedded into everyday work because its products Think about how many times you opened Outlook, Teams, or a Microsoft doc today. Now think about how many times you opened a wellbeing app, or a behavioural coaching tool. Probably zero. Which is exactly why Microsoft's approach is worth paying attention to.
Microsoft has a distribution advantage that no dedicated nudgetech platform can match. Outlook, Teams, Viva, Copilot. Hundreds of millions of people use these tools every day. That reach changes what is possible with a behavioural prompt.
Viva Insights is the clearest example of what Microsoft is doing with it. Let’s say your calendar has been back-to-back for three days and you get a suggestion to block some focus time. Or, a manager on your team has not checked in with someone in two weeks and they receive a gentle reminder.
Microsoft's 2025 Work Trend Index found the employees getting the most out of AI were not just faster, but they were able to prioritise their attention and focus on the most impactful work. Satya Nadella has talked publicly for years about AI extending human capability rather than just replacing tasks. Viva is probably the most concrete product expression of what he means.
3. Perceptyx (inc. Humu) - Timing over information
Humu was founded by Laszlo Bock, the former Chief People Officer at Google who oversaw Projects Oxygen and Aristotle. He had spent years studying what makes teams work. The problem he kept running into was not that managers lacked information. It was that knowing something and doing it consistently are two completely different things.
Most managers understand that recognition matters, that follow-through matters, that making space for quieter voices in a meeting matters. A stressful Tuesday at 4pm has a way of making all of that disappear anyway.
Humu collects HR data like engagement survey results, then uses machine learning to send personalised prompts to managers and employees based on what their specific team needs. Humu’s Nudge Engine increased action in organisations by as much as 250%. When Perceptyx acquired Humu in August 2023, it reflected where the whole employee experience category was heading. Turns out when you send the right prompt at the right moment, people act on it.
4. Workday - HR systems to behaviour
Workday is used by more than 60% of the Fortune 500, which means it sits on more workforce data than almost any other platform on the planet. All that valuable data was typically reviewed by an HR team maybe once a quarter, or after an end-of-year survey. By the time data was analysed the short window for action might have passed.
Peakon Employee Voice challenged that. It uses generative AI to surface themes and emerging issues from employee feedback across 60 languages in real time, then gives managers recommended micro-actions rather than a spreadsheet for them to make sense of. The December 2024 update added rules-based questions that trigger automatically at specific moments in the employee lifecycle, so the feedback is delivered in real-time, rather than waiting for the next survey.
Workday's research found that many managers are already at high risk of burnout, and their generative prompts remove an entire layer of administrative burden from managers’ plates.
5. Thrive Global - AI should acknowledge burnout
In 2007, Arianna Huffington collapsed from exhaustion at her desk, broke her cheekbone, and woke up in a pool of blood. She had been running The Huffington Post and by any conventional measure was succeeding. She just wasn't sleeping, recovering, or functioning well. That experience became the founding logic of Thrive Global, which she launched in 2016 with the premise that burnout is not a personal failing but a systems problem, and that the way to fix it is through small, repeated behavioural changes embedded into everyday work.
Thrive's Microsteps are science-backed prompts that appear inside tools employees are already using, including Microsoft Teams, to interrupt stress, reinforce healthier habits, and embed recovery into the working day. Essentially, a 60-second reset between meetings, or a prompt to set one clear priority before the day starts can really make a difference. The platform has been adopted by more than 200 organisations across 160 countries, including Walmart, where employees using Thrive started seeing measurable reductions in hypertension and pre-diabetes alongside other wellbeing improvements.
Burnout does not just affect how people feel, but also how they think, communicate, and make decisions. A team running on empty makes worse calls, communicates less clearly, and adapts more slowly. Thrive's argument, and the data increasingly backs it up, is that sustainable performance and wellbeing are not separate goals.
6. Cisco - Nudging collaboration
When the pandemic pushed companies towards remote and hybrid working, the informal ways teams stayed connected largely disappeared. This meant managers could no longer tell when someone was struggling just by being in the same room, and disengagement became much harder to spot through a screen.
Cisco heard from clients that team engagement had become their top challenge after that change. Their response was Connected Intelligence, a set of AI agents built into Webex that watch engagement patterns across teams and flag problems early. When someone goes quiet, a manager gets notified. When an internal role opens that fits someone's background, they get a direct nudge to apply. Cisco fills 10,000 internal roles a year, so at that scale catching things early matters.
7. Novartis - Nudging to reduce bias
Most diversity training happens months before a hiring manager sits down with a candidate. By the time the interview starts, most of what was covered has faded.
Novartis took a different approach. Instead of asking managers to remember guidance from a workshop, the company built prompts directly into the hiring process itself. Things like structured interview guides, standardised scorecards and reminders to evaluate on skills and evidence rather than bias.
As Paul Adang, who leads global talent acquisition at Novartis, put it, the aim is to "break down unconscious biases and help see the person behind the stereotype."
It works for the same reason all the other examples in this article work. Guidance delivered in the moment is more likely to change behaviour than when delivered in advance.
8. Salesforce - Nudging coaches
Salesforce's Einstein Coach gives managers prompts based on what is happening across their team in real-time, not what came up in last quarter's review. This could be a reminder to check in with someone before an issue grows, or a nudge to have a one-on-one feedback conversation that's overdue.
Trailhead, Salesforce's learning platform, applies the same logic to learning & development. Employees can earn badges, points, and rankings for completing training modules, embedding skill-building naturally into work rather than a one-off session they sit through once a year. And through their partnership with Thrive Global, wellbeing prompts form part of daily workflows, covering everything from taking breaks to managing stress to protecting focus time.
9. IBM - Tailored nudges
IBM's internal platform Your Learning recommends courses, case studies, and training materials tailored to what each employee is working on. Someone on a cloud computing project gets suggestions relevant to that work, not a generic training catalogue to browse through.
The platform tracks progress and sends reminders to keep going, while managers receive prompts to recognise learning milestones and have career conversations with their teams. IBM also runs internal hackathons so employees can apply new skills in real situations rather than just completing a module and moving on.
IBM has 250,000 employees across 170 countries. Getting the right learning in front of the right person at the right moment is not something you can do manually at that size, which is exactly why they built a system to do it.
10. Amazon - Nudges for safety
Amazon has over 750,000 robots working alongside humans in its warehouses. Keeping those interactions safe at scale is a real operational challenge.
The Robotic Tech Vest was one of their earlier answers. Designed by Amazon Robotics, it uses built-in sensors to alert nearby robots to a worker's presence, causing them to slow down and reroute rather than waiting for a worker to manually flag their location. It had logged over a million activations across 25 sites within its first year.
Amazon has always been a data-driven company. The question they asked is what else that data can do, and safety is one area where nudging behaviour before something goes wrong is considerably cheaper than responding after it does.
11. SAP - Nudges for decisions
SAP SuccessFactors uses AI to reduce inconsistency in the decisions managers make every day. In hiring, the system analyses job descriptions for gender-biased language and suggests alternatives. During performance reviews, AI flags when feedback is drifting toward personality traits rather than measurable outcomes and nudges managers toward more objective language. SAP's AI copilot Joule guides new hires through onboarding, prompting them to complete forms and sign-offs rather than waiting for someone to chase them.
Through Concur, the same logic extends to operational decisions. Employees receive prompts to reduce unnecessary expenses or choose more practical travel options, reducing the cognitive load of making dozens of small decisions throughout the day.
SAP's AI copilot Joule is now embedded across 80% of SAP's most-used business tasks. The bet is that better decisions at that scale add up to something significant.
12. Kraft Heinz - Nudging leaders
Kraft Heinz might not be the first company people associate with workplace AI, but their use of it is pretty fascinating. Partnering with Cultivate, an AI platform that analyses communication patterns across Slack, email, and calendars, they are working towards helping managers become more aware of how they lead.
With employee consent, the AI looks for patterns managers may not notice, whether that is a response rate that has dropped or a team member who keeps getting left out of conversations. When those patterns show up, a manager gets a prompt with a helpful suggestion. For a 100-year-old consumer goods company, this is a big shift in strategy.
Nudge Theory explores how subtle changes in our environment can influence decisions and behaviour without restricting choice. Organisations use nudges to improve outcomes across areas such as learning, wellbeing, productivity, and decision-making.
Based on Nudge Theory by Richard H. Thaler and Cass R. Sunstein (2008), informed by the research of Daniel Kahneman and Amos Tversky. Image Source: McGregor
Trust
The common thread
None of these companies are trying to replace how people work, but to shape the conditions surrounding it in a way that gets the best out of their teams. Google proved that team behaviour is measurable and improvable, Humu showed that timing matters more than content, and Thrive demonstrated that wellbeing and performance are not separate goals.
At the start of this article we noted that peer influence and well-timed prompts change behaviour 3 to 5 times more effectively than formal training alone. These twelve companies are not waiting for AI to get more powerful to act on that. They are already using nudgetech to influence how people communicate, collaborate, learn, and lead, and the results in engagement, retention, safety, hiring, and wellbeing data speak for themselves.
The technology is not the hard part. The challenge is building enough trust among teams to get the most out of these new tools.
