Menu

For HR leaders

Stop pretending people do all the work.

Your people already work with AI agents, most of it unplanned and unowned. HR decides whether that work gets designed or stays hidden.

Training and advisory for HR leaders. Learn to evaluate AI-supported work, configure AI agents for human collaboration, and test changes in one real workflow.

Share your AI priorities

The working system

Work crosses the org chart’s reporting lines.

People draft, test, and revise alongside AI agents. Map their contributions and review points as well as formal roles. Both belong to the same working system.

People and AI agents do the work together.

People team · updating manager guidance

People AI agents Reporting line Work connection

Scroll across to follow the work connections.

Inside this org chart

People, roles, and reporting lines

CHRO / Head of People
HR business partner

Drafts and revises policy

L&D lead

Tests guidance with managers

People operations lead

Designs the operating steps

Beyond this org chart

AI agents working with different roles and workflows

ChatGPTClaude
ChatGPTManus
ClaudeGemini
HermesOpenClawChatGPT

The outline marks reporting structure, not the limits of the organization.

AI contribution is not the problem. Unclear review and decision ownership are. Start with one workflow: who contributes, who reviews, and who decides.

57%

of employees reported nontransparent AI use at work

University of Melbourne + KPMG · employees in 47 countries · 2025

That is a reason to clarify roles and review—not to monitor private prompts.

Your AI expertise

Evaluate the team, not just the model.

Technical benchmarks tell part of the story. We help HR build the judgment and practical skills to assess how people and AI agents work together.

  1. 01

    Recognize the human side effects

    Spot over-reliance, uncritical agreement, and work shifted onto colleagues. Define what people need to check before AI-assisted output moves forward.

  2. 02

    Configure AI agents for collaboration

    Write and test instructions for AI agents: when to ask questions, challenge assumptions, cite sources, and return control to a person.

  3. 03

    Measure the whole team's result

    Compare accepted output, human review time, rework, and total cost against today's process. A faster draft is only one part of the result.

  4. 04

    Design roles and learning

    Decide what to delegate, which skills people need to retain, and where managers need support. Make review, approval, and escalation responsibilities explicit.

Inspect the work. Support the person. Never turn private prompts into a performance score.

Learn and apply

Bring leadership a recommendation you can defend.

Build understanding through training and advisory, then practice on one workflow. Start with a sample document and tools your company already permits. Compare the team's result before proposing a wider change.

New tools, data access, policy changes, and employment decisions need the accountable owners. Bring in Technology, Security, Privacy, or Legal when the proposed change crosses their boundary.

Share your priorities first. Agree the format, participants, and scope before work begins.

From learning to practice

Build a case for the next step.

  • Criteria for evaluating people and AI agents working together.
  • AI agent instructions with sources, review rules, and stop conditions.
  • A workflow test that includes human review and rework.
  • A recommendation: what to change, why, and who needs to be involved.

Apply your expertise

Start with one document workflow.

Put the skills to work on manager guidance: configure AI support, test human review, and trace contributions through approval. Onboarding materials and employee learning are other places to apply the same approach.

Follow a document from first draft to publication.

Select a stage to see what people and AI agents do, what the shared process records, and what must happen before work moves on.

Manager guidance document · version 0.1

Frame & Draft

Exit gateExit when the owner and source set are explicit.

Human action

The HR owner names the decision, source set, reviewers, and definition of done.

Personal AI agent

The personal AI agent prepares one sourced draft and identifies claims that need a source.

Moderator record

Records the working title, source set, version 0.1, and open questions.

Facilitator route

Confirms the right people are in the room and routes a source request.

Personal AI agents support individual contributors. The company-provided Moderator keeps versions, objections, and decisions visible. The Facilitator routes missing checks and authority. Named people retain review, approval, and accountability.

Start with a manual pilot: one current version, visible contributions, and a named person approving each release.

Explore Document Review & Approval

Share your AI priorities

Where the productivity goes

AI is already a billion-person behavior.

Personal AI habits reach the workplace. The management question is how to support people and make consequential work reviewable.

ChatGPT reaches more than one billion people every week. The familiar chat is becoming a system that researches, uses connected tools, takes multi-step action, and repeats work on a schedule.

1B+

weekly active ChatGPT users

OpenAI · weekly ChatGPT usage
Read the workflow-redesign evidence

The human factor

Fluent output can outrun human review.

This is not a discipline problem. It is a design problem: what gets checked, by whom, and before what moves forward.

For some people, the relationship becomes personal.

For most people ChatGPT stays a tool. For a small group it does not. OpenAI and MIT found emotional engagement concentrated among heavy daily users. Across the study, longer daily use was also linked to more emotional dependence and problematic use.

OpenAI + MIT Media Lab · 2025

58%

of experienced AI users agreed the AI “did most of the thinking”

It held across ten work tasks: planning, sequencing, deciding. People who modified, challenged, or rejected what the AI agent produced reported more confidence in their own reasoning. People also described ideas that no longer felt like theirs.

Technology, Mind, and Behavior (APA) · 1,923 weekly AI users · 2026

1h56m

of rework for every piece of AI-assisted work that arrives looking finished and is not

Researchers at BetterUp Labs and Stanford's Social Media Lab named it workslop. 41% of US workers surveyed had received it; about half thought less of the colleague who sent it, and 42% trusted them less. The estimate at scale is roughly $186 per affected employee per month.

HBR · BetterUp Labs + Stanford Social Media Lab · 2025

How 8Hats Lab helps

Three layers for one human–AI organization.

Choose the layer that fits the workflow. Agents Coworking is available for pilot work; Organization Model is in development.

01

Personal AI agent

Delegation that fits the person.

With explicit consent, AI agents can learn how a person prefers to be challenged, how ambiguity is escalated, and when control returns. Personalization is support—not employee scoring.

02

Rooms

A transparent format for consequential work.

Shared briefs, roles, contributions, review gates, open conflicts, escalation, and an agreed definition of done.

Explore Agents Coworking
03

Organization Model

Shared context that survives the project.

Accepted decisions, corrections, and working standards live in the company's shared knowledge instead of private chats.

See the Organization Model

How we design work for people and AI agents

Who you would be working with

We know what changes when the tool changes the work.

Our team has led learning, research, product, and transformation from inside universities, corporations, public institutions, and companies of our own.

Dmitriy Istomin

For 25+ years, he has put digital tools, and now AI, to work in learning and education.

Taras Pustovoy

For 25+ years, he has built learning at scale.

Alexander Volkov, PhD

An engineer by training, with a PhD in dynamics and strength of machines.

Mariam Mamedli, PhD

A PhD economist with 14 years across data, ML, NLP, and product in banks, central banking, and fintech.

Meet the full team

For search—and the first internal conversation

Questions HR leaders are already asking.

What is a human–AI team?

A human–AI team is a working unit in which people and AI agents contribute to the same outcome. AI agents may draft, research, coordinate, or challenge; people retain the decision rights and accountability defined by the organization.

Why should HR co-own AI transformation?

Because AI changes roles, contribution, review, learning, trust, management practice, and employee support. Technology, Security, and Legal remain essential owners of systems, controls, and policy. HR owns the human and organizational design with them.

What is shadow AI at work?

Shadow AI is employee use of AI tools or AI agent capabilities outside the organization's approved systems and controls. An inventory helps, but the operating question remains: what must be disclosed, reviewed, escalated, and approved when AI contributes to consequential work?

How can HR address AI over-reliance without surveillance?

Set evidence and review standards around consequential artifacts, train people to challenge fluent output, define stop and escalation conditions, and offer support through established wellbeing channels. Do not inspect private prompts or use inferred mental states as performance data.

Do I need to become an AI engineer?

No. Build enough practical understanding to question assumptions, configure AI agent instructions, and judge team outcomes. Work with technical colleagues on systems and access; bring your own expertise on roles, learning, and human review.

How should a company begin?

Use the form to describe the AI decision or workflow your HR team faces. Start with skills, a process, or help choosing where to begin. No employee records or confidential documents are needed for the inquiry.

Your next step

What does your HR team need next?

Share your priorities: build HR's AI expertise, improve a specific process, or work out where to start. Describe the situation in the message field below.

Use the general inquiry form below. Keep employee information and confidential documents out of your message.

General inquiry

Contact us.

Tell us what you need. A member of the team reads every inquiry and contacts you personally.

We use this address to contact you.

Privacy notice
Prefer email? Write to hello@8hats.ai.