# AI Takes Over at Goldman Sachs as Jobs Vanish for Good

> Goldman's AI-driven cuts mark an industry reset. HR must redesign work around AI-map tasks, set standards, redeploy talent, and measure ROI to cut risk and lift performance.

- Type: AI news
- Page: https://completeaitraining.com/news/ai-takes-over-at-goldman-sachs-as-jobs-vanish-for-good/
- Published: 2025-10-18
- Updated: 2025-10-18
- Author: Jeroen Erne
- Jobs: Human resources
- Reading time: 3 min

### AI Cuts At Goldman Signal A New Baseline For HR

Goldman Sachs plans to reduce roles where AI now outperforms people. Many who lose those roles will not get them back. This is not a temporary cycle. It's an industry reset.

If you lead HR, treat this as a blueprint moment. Rebuild job design, skills, and workforce planning around AI as a core tool, not a side project.

#### What This Means For HR Leaders

- **Job recovery is unlikely** for tasks that AI can produce faster, cheaper, and with consistent quality.
- **The shift is task-first**: AI replaces task bundles inside roles before it replaces entire roles.
- **Winners reallocate talent** to higher-value work instead of only cutting headcount.
- **Policy, compliance, and skills** will decide whether your AI program saves costs without creating new risk.

#### Immediate Actions (Next 30-90 Days)

- **Map tasks, not titles:** For each role, tag tasks as Automate (A), Assist (S), or Human-only (H). Prioritize high-volume A/S tasks for redesign.
- **Freeze and redesign:** Pause hiring on roles with 30%+ A-tasks. Repost as hybrid roles with clear AI expectations and output targets.
- **Set standards:** Write "AI-use clauses" into job descriptions, performance reviews, and SOPs. Define who is accountable for final outputs.
- **Stand up an AI review board:** HR, Legal, Risk, Data, and Procurement. Approve tools, vendors, and monitoring. Log decisions.
- **Baseline AI literacy:** Run short, role-based training and practice labs. Start with office suites, analysis, and prompt skills. [See AI courses by job](https://completeaitraining.com/courses-by-job/).

#### Redeploy, Don't Just Reduce

- **Internal mobility first:** List A/S task owners at risk and match them to revenue, risk, or client work that needs headcount.
- **Skill sprints:** 4-6 week micro-credentials for data analysis, workflow automation, and compliance documentation.
- **Incentives:** Tie redeployment bonuses to completion of projects that show time saved or error-rate cuts.
- **Measure ROI:** Compare severance cost vs. reskilling plus time-to-productivity. Fund the option with higher long-term value.

#### Role Redesign Examples (Banking Context)

- **Analyst:** From building slides and models to validating AI outputs, testing scenarios, and client-ready synthesis.
- **Compliance ops:** From manual reviews to supervising AI triage, exception handling, and audit trails.
- **Research associate:** From raw data pulls to query design, fact-checking, source ranking, and risk notes.
- **New roles:** Automation operations lead, prompt and workflow specialist, AI risk analyst.

#### Policies, Risk, and Compliance

- **Bias and impact checks:** Test outputs for disparate impact. Document fixes and retests. See [EEOC AI guidance](https://www.eeoc.gov/laws/guidance/select-issues-assessing-adverse-impact-software-algorithms-and-artificial-intelligence?ref=completeaitraining.com).
- **Data controls:** Block sensitive data from public tools. Use approved vendors. Require logs and explainability for material decisions.
- **Auditability:** Keep prompts, outputs, and human approvals for regulated processes.
- **Vendor contracts:** Include SLAs for accuracy, security, incident response, and model change notices.

#### Hiring For An AI-First Team

- **What to screen:** systems thinking, data literacy, prompt writing, judgment under uncertainty, and basic scripting familiarity.
- **Practical tests:** Give a real workflow, an AI tool, and 60 minutes. Judge output quality, verification steps, and documentation.
- **New expectations:** Every role ships more output per FTE with AI. Performance plans should show how.

#### Metrics That Matter

- Percent of tasks by role tagged A/S/H (and updated quarterly)
- Time-to-output and error rate before vs. after AI
- Internal mobility rate for at-risk employees
- Training completion and tool adoption by team
- Incidents: bias, data leakage, policy exceptions
- Cost per deliverable vs. baseline

#### Communication Playbook

- **Be explicit:** Share the task map, selection criteria, and timelines. No vague promises.
- **Offer paths:** Provide clear reskilling tracks with guaranteed interviews for target roles.
- **Support exits:** Strong severance, coaching, and alumni referrals. Treat people well and protect your brand.

#### Context You Can Use With Leadership

- Large banks will automate a meaningful slice of research, operations, compliance, and support tasks.
- External estimates indicate broad task exposure to AI across industries. See [Goldman Sachs research](https://www.goldmansachs.com/intelligence/pages/generative-ai-could-raise-global-gdp-by-7-percent.html?ref=completeaitraining.com).

#### The Bottom Line

AI-driven cuts at Goldman are a preview, not an outlier. Some roles will not return. Your job is to reduce risk, move talent to higher-value work, and set new performance standards.

Move quickly, measure everything, and rebuild roles around AI from the task level up.

---

This is the Markdown form of https://completeaitraining.com/news/ai-takes-over-at-goldman-sachs-as-jobs-vanish-for-good/. The whole site is mapped at https://completeaitraining.com/llms.txt.
