Happy Thursday, everyone. I'm Frank Richardson, an organisational psychologist observing the workplace with curiosity and care. Each week, I share insights to help HR leaders better understand the people behind the processes and build cultures where both individuals and organisations can thrive.
This week in workplace whiplash 🌀
A few stories this week that quietly reveal what’s shifting beneath the surface at work:
🤖 Google’s AI boss sounds the alarm
The head of Google DeepMind said urgent research is needed to tackle AI threats, warning that as systems grow more powerful, risks related to biosecurity, cybersecurity and misuse by malicious actors are becoming more pronounced. Suggesting that safety impacts of AI are now too big for individual companies to manage alone.
👉BBC👥 “Job hugging” might be quietly killing impact
Research from MetLife suggests that employees who cling to roles, not out of engagement but fear, are less productive and less innovative than their more adaptive peers. Staying put doesn’t always mean loyalty; sometimes it means avoidance.
👉HR Dive💸 Workers are losing hope on wage growth
Many employees believe wage increases aren’t keeping pace with expectations or cost-of-living pressures, leading to declining confidence in future earning potential. When hope around pay erodes, motivation often drains slowly, not suddenly, which makes long-term retention even harder.
👉SHRM
And on the topic of trends that make work feel harder and messier… let’s talk about what happens when AI removes friction from workplace complaints.
Many years ago, I filed a formal workplace complaint.
I rewrote it obsessively. I checked policy wording. I deleted half of it because I worried I sounded dramatic. It took weeks. I lost sleep. Submitting it felt like pressing a big red button.
There was friction. And the friction mattered.
Which is why I nearly choked on my coffee reading a recent Financial Times piece about HR teams now receiving beautifully structured, legally fluent, AI-assisted grievances, generated in minutes. Perfect tone. Impressive formatting. Occasional hallucinated legislation.
The issue is, when escalation becomes effortless, HR doesn’t just get more complaints, it gets completely overwhelmed. And overwhelmed systems don’t make sharper judgements... they make faster ones.
🧠The behavioural science lens
Once you strip away the AI hype, this becomes a story about how humans judge under pressure:
Authority bias kicks in quickly: Research on authority bias shows how strongly we respond to signals of institutional credibility. When a complaint reads like it was drafted by a barrister, cites legislation and uses confident legal phrasing, it activates authority cues. Even when we know AI may have written it, the tone itself nudges us toward deference.
Fluency makes things feel true: A classic study found that statements presented in an easier-to-read format were judged as more likely to be true, even when identical in content. AI-generated text is extremely fluent and our brains often mistake that smoothness for strength.
Polish substitutes for proof: The Elaboration Likelihood Model shows that under pressure, we rely on peripheral cues like confidence, structure and authority rather than carefully evaluating arguments. A 20-page, well-formatted grievance feels serious… even if the substance is patchy.
Cognitive load lowers scrutiny: As summarised in Thinking, Fast and Slow, when we’re overloaded, we default to shortcuts. HR teams reviewing dense AI-generated complaints at volume are operating under strain, and strain increases reliance on heuristics.
🚀What this means for leaders
If anyone can produce a barrister-level grievance in ten minutes, your systems may need to level up.
Stop equating length with legitimacy: More words do not equal more merit. Build structured intake frameworks that force clarity instead of rewarding verbosity.
Train HR on fluency bias: Polish feels powerful. It isn’t proof. Decision-makers need guardrails that separate tone from evidence.
Protect bandwidth like it’s governance: Overloaded teams make inconsistent decisions. Inconsistent decisions create legal risk. While it's an admin issue, it’s also a huge risk issue.
Increase procedural transparency: When escalation becomes easier, clarity becomes non-negotiable. Employees can tolerate tough outcomes. They don’t tolerate opaque ones.
Don’t panic about AI, instead fix the system: Some employees will use AI to articulate genuine concerns more clearly. The real question isn’t whether they should. It’s whether your decision-making architecture is strong enough to handle it.
💬 Final thoughts
When I wrote that complaint years ago, the effort forced me to think carefully about what I was alleging.
AI removes that effort, and while it doesn’t automatically make complaints worse, it does mean the hard work now sits with the people assessing them.
And if we don’t strengthen how we judge, we’ll mistake smoothness for substance.
How's the depth of today's edition?
If something here speaks to you, I’d love to hear it.
Until next week,
Frank
P.S. If you want to get a feature about your own story, reply to this email. If you’d like to reach our newsletter audience (founders, creators, and marketers), click the button below.
If you’re new here, I’m over the moon you’ve joined us! To help me craft content that’s actually useful (and not just noise in your inbox), I’d love it if you took 1 minute to answer this quick survey below. Your insights help shape everything I write.
✨ Insane Media is more than one voice
💡 Dive into our other newsletters - where psychology meets the founders, creator economy, e-commerce marketing, and AI founders.
