AI has crossed the experimental phase in contract lifecycle management (CLM). Gartner projects that 50 % of all procurement contract writing tasks will be AI‑enabled by 2027 , while World Commerce & Contracting sees AI adoption as the fastest lever to recover the 8‑9 % of value that typically “leaks” from supplier agreements . Matt Lhoumeau’s July 2025 article in Procurement Magazine crystallizes the moment: by 2035, routine legal work around buy‑side contracts will be executed largely by autonomous AI agents .
By Per‑Erik, guest lecturer at Learn How to Source – July 25 2025
Content…
Why this conversation matters now
Two days ago Matt Lhoumeau (CEO of CLM‑vendor Concord) dropped a grenade in Procurement Magazine: by 2035, routine legal co‑workers will be “obsolete,” displaced by AI agents that draft, risk‑score and even approve most supplier agreements. (procurementmag.com)
Bold? Absolutely. But Matt isn’t alone: World Commerce & Contracting’s annual survey shows contract professionals already piloting AI to recover the 8.6 % of contract value that normally “leaks” during a term; their top hope for 2025‑27 is value realisation, not just speed. (info.worldcc.com)
This post unpacks what that means for early‑career buyers who today lean on a trusty Word template attached to every RFQ:
- What AI can already do for contract drafting, review and negotiation.
- How tomorrow’s “dynamic templates” and agentic AI could re‑wire the workflow.
- Practical guardrails so you stay in control while the machines take the busy‑work.
Where we are today: template‑driven contracting
Most sourcing teams still follow a familiar pattern:
- Draft bespoke template. Legal edits last year’s Word doc to fit the new category scope.
- Attach to RFQ. Suppliers accept (or red‑line) as one of several bid documents.
- Manual review & redline. Each counter‑party mark‑up kicks off an email ping‑pong.
- Execute & archive. Signed PDF disappears into a shared drive until something goes wrong.
It’s workable, but slow. Deloitte’s 2025 legal survey found that cycle times for mid‑complexity buy‑side contracts still average 30–45 days. (Deloitte Italia)
What AI already changes in 2025
| Contract phase | Real‑world AI assist (2025) | Tools on the street |
|---|---|---|
| Pre‑draft | Suggests optimal template and clause set based on category, spend, Incoterms and governing law. | Ironclad “Global Clause Library”, ContractPodAI Leah recommends template on intake. (ContractPodAi) |
| Drafting | Generates first‑draft contract from RFQ metadata; fills commercial tables and embeds standard Ts&Cs. | Concord “Instant Draft,” Spellbook add‑in for Word. (concord.app) |
| Review | Parses third‑party paper; flags non‑standard language, scores risk, proposes fallback clauses. | SpotDraft AI review, Henchman, Thomson Reuters CoCounsel. (LawSites) |
| Negotiation | Suggests compromise language and tracks red‑line concessions vs. playbook in real time. | Market‑wide: DocuSign CLM AI, Icertis NegotiateAI. |
| Execution & post‑sign | Validates signature blocks, extracts obligations, sets renewal alerts, monitors KPI compliance. | Agiloft Obligation AI, Ironclad “Entities & Obligation Management”. |
Gartner says that by 2027, half of all procurement‑contract‑management steps will be AI‑enabled—up from <10 % three years ago. (concord.app)
What happens to the humble template?
Spoiler: it doesn’t die – it evolves.
- From static to dynamic. Templates become parameter‑driven blueprints that an LLM assembles on the fly. You supply five data points (scope, jurisdiction, payment model, risk tier, ESG clauses) and the engine pulls the right boilerplate.
- Clause‑level learning. Each negotiation trains the model: if suppliers always push back on an IP‑indemnity line, the AI suggests an alternate clause or pre‑approved fallback next time.
- Auto‑sync with sourcing data. Win‑prices, SLA metrics and supplier risk scores flow straight from your eRFx platform into the contract header – zero re‑keying.
- Continuous compliance. When EU Forced‑Labour regulation hits, the system bulk‑amends hundreds of live contracts and routes only the high‑risk ones to human legal for approval.
WorldCC’s 2025 white paper calls this “agentic architecture”: a network of narrow AI agents collaborating on clause selection, risk checks and performance monitoring. (worldcc.com)
2030‑plus: autonomous agents in the loop
ContractPodAI predicts that by 2028 at least 15 % of day‑to‑day contract writing and connected decisions (e.g., choosing a fallback clause) will be made autonomously by agentic AI cohorts. (ContractPodAi)
McKinsey’s “Superagency” research argues the same pattern across business workflows: humans set guardrails; AI handles repetition and reasoning at scale. (McKinsey & Company)
For procurement, that means:
- Self‑service contract writing and contracting for < €100 k PO’s: requestor picks a drop‑down scope; AI drafts, risk‑scores and e‑signs in minutes.
- Continuous negotiations: an AI bot monitors commodity indices and auto‑triggers price‑adjustment clauses.
- Smart obligations feeding back to supplier‑performance dashboards, closing the loop between contract promises and category KPIs.
But buyers still matter – here’s how
| Human value‑add | Why AI won’t replace it (yet) | Your up‑skilling to‑do |
|---|---|---|
| Commercial creativity | AI can remix past clauses, not invent net‑new partnership models. | Practice prompt‑engineering to push AI beyond boilerplate. |
| Stakeholder diplomacy | Conflict resolution, culture fit and political nuance stay human. | Hone negotiation soft skills & cross‑cultural EQ. |
| Ethical & regulatory judgment | AI may hallucinate or miss future‑dated regs. | Learn AI‑governance frameworks; set guardrails. |
| Category strategy linkage | Deciding when to dual‑source vs. single‑source is a business‑risk call. | Strengthen analytical thinking and scenario planning. |
SpotDraft’s 2025 survey shows teams that use AI daily report 91 % productivity uptick, but still rate strategic thinking as their #1 future skill. (LawSites)
A first‑step roadmap for early‑career buyers
- Audit your template library. Tag each clause with risk tier and fallback; clean data is AI fuel.
- Pilot “AI review” on NDAs. Low‑risk, high‑volume, clear payback.
- Integrate RFQ metadata. Map the eight fields your eSourcing tool can pass straight into the draft contract.
- Set up a dual approval lane. AI auto‑approves green‑score contracts; amber/red goto legal.
- Track cycle‑time KPIs. Show hard savings (hours + legal fees) to win budget for bigger roll‑outs.
Remember: adoption stalls when users don’t trust the outputs. Tie every pilot to a human‑in‑the‑loop checkpoint so confidence grows with each successful event.
Reading list – dive deeper into AI‑driven contracting and contract writing
| # | Topic | Source (HTML) | Year |
|---|---|---|---|
| 1 | Provocative future view | Why Procurement’s Legal Coworkers Will Be Obsolete by 2035 – Procurement Magazine – https://procurementmag.com/news/procurements-legal-coworkers-obsolete-2035 | 2025 |
| 2 | Adoption data & value‑leak numbers | World Commerce & Contracting – AI Adoption in Contracting – https://info.worldcc.com/ai-adoption-in-contracting | 2025 |
| 3 | Analyst prediction (50 % AI‑enabled by 2027) | Concord blog – AI for Contracts – https://www.concord.app/ai-for-contracts/ | 2025 |
| 4 | Legal‑department outlook & five predictions | Deloitte – AI for In‑House Legal: 2025 Predictions – https://www.deloitte.com/global/en/services/legal/research/ai-inhouse-legal-2025-predictions.html | 2025 |
| 5 | In‑house caution & skills survey | LawNext – SpotDraft AI Impact on Legal Teams – https://www.lawnext.com/2025/04/in-house-counsel-embracing-ai-for-contracts-but-cautiously-finds-new-survey-from-spotdraft.html | 2025 |
| 6 | Agentic AI & CLM trend deep‑dive | ContractPodAI – Why 2025 Demands AI‑First Strategies for CLM – https://contractpodai.com/news/clm-ai-strategies/ | 2025 |
| 7 | “Superagency” concept & workforce impact | McKinsey – Superagency in the Workplace: Empowering People to Unlock AI’s Full Potential– https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/superagency-in-the-workplace-empowering-people-to-unlock-ais-full-potential-at-work | 2025 |
Bottom line: templates aren’t disappearing—they’re getting smarter. Your job is to feed the machine good data, keep one hand on the strategic wheel, and let AI handle the drudge work of contract writing. Start small, learn fast, and by the time 2035 arrives you’ll be running a sourcing function where legal “obstacles” are solved before the first draft even lands in your inbox.
Have a look at Learn How to Sources course about Sourcing Engine and learn about LHTS standard Sourcing process and the course about creating AI agent in Chat GPT. Learn where contract writing fits in the sourcing process.
Happy contracting!
— Per‑Erik
Inline citations: Gartner, WorldCC, Deloitte, SpotDraft and others embedded above. Extract and illustration to blogpost about Contract Writing was done with support from Chat GPT o3.