FORWARDED THIS? Subscribe here to get the next one.
ISSUE 06 · AI + COMPENSATION
ChatGPT now hands back finished comp work
TL;DR
OpenAI merged ChatGPT and Codex into one app that hands back finished spreadsheets, decks and dashboards. We wrote the comp guide for it, your poll picked the theme, and the radar has a pay transparency deadline landing on Tuesday.
A NOTE FROM GIAC
Quick one this week. On Monday I asked the community what pushback they hit most when pitching AI in Rewards, and the answer came back louder than I expected. Legal and data privacy, by a mile.
The same week, I kept coming back to what OpenAI shipped on 9 July. An agent that hands back finished work sounds like the opposite of what a privacy worried legal team wants to hear. It doesn't have to be. The trick is deciding what the agent gets to touch, and this issue is built around exactly that.
Giac
THE BREAKDOWN
OpenAI's agent does the work now. The judgement is still yours.
On 9 July OpenAI merged ChatGPT and Codex into one desktop app. The agentic half is called ChatGPT Work. You give it an outcome, it works on its own for minutes or sometimes hours, and it hands back a finished piece of work. A spreadsheet, a slide deck, or a live dashboard you can open and share with a link.
It runs on every plan, including the free one. That part matters for comp. The capability that used to sit with a data team now sits on your laptop, and the people you support can reach it too.
Two weeks in, the comp use cases are real. Merit cycle dashboards, survey cleanup, levelling passes, pay equity scripts you can hand to an auditor. The judgement calls haven't moved an inch. It still doesn't know your aging factors or whether your regression controls for the right things, and it sounds confident either way.
Because it's an agent with file access and connectors, the data question comes before everything else. The guide leads with the rule that answers it. Build on dummy data, then run the finished thing against real pay data yourself, on your own machine.
The full guide covers the seven use cases, the five places your rigour still carries the work, which plan you actually need, and what to do instead if your company runs on Microsoft.
FROM RANGE
Six of you went public this week, and it means a lot to see people put their name to building comp with AI. If any of these faces are new to you, they're worth following.
The wall you hit most, in your own words
Monday's poll inside the community asked what pushback you hit most when pitching AI in Rewards. Legal and data privacy won by a landslide, ahead of budget, ROI proof and leadership scepticism combined. A working group on handling that conversation is taking shape inside Range now.
One member flipped the whole question. Their leadership is all in on AI, which brings its own pressure when you work with the most sensitive data in the company. They said the tool access is part of what keeps them in the job, and they'd take that over fighting for it any day. If AI enablement is becoming part of the employee value proposition, retention conversations are about to get interesting.
Your turn below. Same question, whole list this time.
What's the pushback you hit most when pitching AI in Rewards?
ON THE RADAR
Amazon is cutting AI roles while spending 200 billion dollars on AI. On 22 July it confirmed layoffs inside the AGI team building its frontier models, even as it ramps record spending on AI infrastructure. Its CEO has said in plain terms that generative AI will shrink the corporate workforce. The clearest sign yet that AI budget and headcount are now one conversation. Read → the cuts
Meta is being sued by 26 employees over AI assisted layoff selection. Filed 13 July in an Oakland federal court, with claims running through family leave, disability and pregnancy discrimination law, plus a claim the AI was never tested for bias. The first major challenge to AI assisted layoff decisions, and the models in question scored people on productivity data. Read → the case
Starbucks just tied store pay to a quarterly scorecard. Rolled out across its US stores this month. Hourly staff earn 300 dollars a quarter when their store hits its sales, operations and service targets, up to 1,200 a year, and with new tipping options the company puts the total lift at roughly 5 to 8% of pay. A template worth watching for frontline variable pay. Read → the programme details
FROM THE COMMUNITY
The privacy working group forming off the back of Monday's poll is the community doing what it's for. Somewhere to pressure test a build before it touches a real payslip, with people who've already had the conversation your legal team is about to start.
RECENT EDITIONS
Reading this because a friend forwarded it? Get Range Weekly in your own inbox every Sunday. Subscribe free |
 That's me for this week. What would you point an agent at first, and what would you never let it near? Hit reply and tell me, I read every one. Until next week, Giac |
 |
When you're ready, here are three ways Range can help you.
1. Join the community
Find your people. Inside Range, reward pros learn to build with AI together, the prompts, the setups, the band workflows that actually hold up. Free is the signal, the room is the substance. Apply to join →
2. Grab the free resources
Everything free we've made now lives in one place, the Comp AI Vault. The guides, the prompt library and the frameworks, all in the open. Browse the Vault →
3. Reward Rewired 2026
Our in person gathering for the people actually building this in comp. Register your interest →