The HC Commodities Podcast – Episode Summary

Episode Overview

Episode Title: Techno-Barbarians at the Gate: AI is coming for your operating model
Host: Paul Chapman (HC Group)
Guest: Eren Zekioglu (former Glencore and Gunvor COO/CIO)
Date: February 18, 2026

This episode explores how artificial intelligence (AI) is transforming the operating models of commodity trading firms, its current state and future trajectory, and the interplay between AI, digital assets, and decentralized finance (DeFi). Eren Zekioglu leverages his deep operational and technological experience in hedge funds and trading houses, delivering nuanced insights on adoption challenges, organizational culture, compliance, and the evolving demands of talent in the commodities sector.

Key Discussion Points & Insights

1. AI in Commodities: Already Here, Not Fully Transformative (02:40–05:33)

• AI is not the future; it’s already present in the sector—used in forecasting, trade idea generation, process automation, compliance, and document management.
• Current impacts are primarily surface-level: incremental process optimization and efficiency, rather than radical transformation.
• The anticipated “big reduction” in headcount or reconfigurations hasn’t occurred yet due to early-stage adoption and significant legal/compliance risk.

“When I hear the conversation about AI coming into commodities, I'm like, well, AI is actually already here ... we're using it for more optimization and improving our processes across a surface.”
— Eren Zekioglu (02:43)

• The field is split into two broad AI buckets:
  • Alpha generation: Better trading decisions and automated idea creation.
  • Process efficiency: Agentic AI that reduces manual workflow, mainly in operations.

"Within the commodity space it's still very, very early. There's not enough people that really understand the full capabilities of it..."
— Eren Zekioglu (04:30)

2. Comparing Organizational Adoption: Hedge Funds vs. Trading Houses vs. Asset-Backed Majors (06:44–10:59)

• Hedge funds are ‘structurally ahead’:
  • Built on models, faster deployment, cleaner data, and more ‘web3-literate’ culture.
  • Quick to experiment and adopt working AI.
• Trading houses:
  • More complex due to legacy systems, cultural resistance, multiple process layers, and both internal/external stakeholders.
  • Slower to adopt due to necessity for “blessings and baptizing” from legal, compliance, and the trader themself.
• NOCs (National Oil Companies) & Asset-Backed Majors:
  • Potential leapfroggers in AI adoption because they're starting with a blank slate, can implement modern stacks from day one, and face fewer entrenched processes.

"Trading houses are far more complex… trying to implement AI internally also has to be a little bit externally as well because we have so many people involved."
— Eren Zekioglu (08:04)

3. Known Applications: Where AI Is Delivering Today (13:03–17:47)

• Trading/Front Office:
  • Accelerated trade decision generation, faster forecasts, macro and market analysis.
• Middle/Back Office:
  • AI-driven document processing, compliance monitoring (KYC/AML), and error reduction.
  • Potential for agentic AIs to collapse middle/back office into a single streamlined unit.
  • Shipping/logistics optimization benefits.
• Compliance/Surveillance:
  • Massively increased speed and comprehensiveness of compliance muscle; now approaching real-time due diligence.

"AI has definitely reduced our decision time, our execution time, our settlement time, and especially the time it takes to deploy capital, which is enormously important for us."
— Eren Zekioglu (16:50)

4. Limits of AI Today: The Importance of Proprietary Data and Human Judgment (19:18–23:45)

• Tradable advantage from proprietary data is eroding; AI enables faster processing but doesn’t create unique data itself.
• Paper/electronic trading: AI can suggest and execute trades flawlessly.
• Physical trading: Still opaque, highly relationship- and trust-based; AI complements but cannot fully replace the human trader’s role.
• "Guild work" (rote, rules-based tasks) is most threatened by AI.

"For physical trading, I think the human trader still needs to make the decision... for a paper trader, yeah, honestly, I just think you won't need historical old baraboy traders anymore. I think you'll need much more computer science traders for sure."
— Eren Zekioglu (21:35)

5. The Collapsing Office Structure and De-Siloing (23:45–25:29)

• Prediction: Middle/back office functions will vanish or merge into a single ‘office’—eventually, these distinctions and even "AI" as a term will disappear and become part of the background infrastructure.
• NOCs and new entrants can move directly to this AI-native model, leaving legacy players at risk of being left behind or outpaced.

"I don't think there'll be a middle and back office anymore. It's just going to be an office … AI won't even be a word. It will just be normal."
— Eren Zekioglu (23:53)

6. Unknowns & Transformative Potential: AI, DeFi, and Digital Assets (26:29–33:34)

• AI as a driver for new entrants: Tech and stablecoin firms are launching trading arms, enabled by 24/7 AI/DeFi operating models.
• Integration of trade, finance, and logistics possible in single, AI/DeFi-driven platforms.
• Asset-backed majors and NOCs (already leaders in AI-driven drilling, etc.) are becoming trading powerhouses—information footprint and operational agility give them an edge.
• Possibility: Merging of industries between tech and commodity players; new ‘hybrid’ entities feasible.

"AI has given us confidence … the capabilities now are endless. We can be a player across the life cycle ..."
— Eren Zekioglu (30:06)

"There are various commodity traders that have digital asset desks ... for them it's an asset which is making money. So why shouldn't they hold it as part of their treasury?"
— Eren Zekioglu (32:24)

7. People, Culture, and Leadership for the AI Future (34:13–45:10)

• Transformation must be led from the top (C-suite):
  • C-suite must be fluent in digital assets, AI, and DeFi.
  • Otherwise, the new CIO or CTO will waste years educating leadership and risk costly dead-ends.
• Talent needs: Demand is shifting rapidly towards traders and executives with digital assets, AI, and tech fluency.
  • Commodities industry knowledge can be taught, but digital sophistication is now a rare and vital asset.
• Chief Operating Officers, CFOs, risk, compliance, and legal: All roles now require AI and digital asset experience.

"You need to start at the top. Now if there is an absence of this knowledge, which there probably is at the sort of CFO/COO level, then obviously you call the consultants in…"
— Eren Zekioglu (37:00)

• AI is the “great democratizer”: Employees already use it in daily life; organizations must catch up or risk obsolescence.

8. AI is Here to Stay, but Judgment and Relationships Remain Irreplaceable (43:09–45:10)

• AI will automate what can be standardized but enhance—rather than replace—the high-value work of judgment, relationships, and major strategic calls.
• The operating model of the future is digital asset-native, DeFi-ready, and led by the curious and tech-literate.

"AI will be adopted, but you still need the physical trader. You will still need them. It's not going to wipe that out."
— Eren Zekioglu (44:04)

Notable Quotes & Moments

• On AI’s real-world adoption:
  "Anyone can cut, paste a moving average trade and stick it into a generative AI tool and it will give it some various long and short options ... Document processing has, has increasingly helped with AI, the compliance monitoring... AI has completely changed that operating model."
  — Eren Zekioglu (14:17)

• On talent and leadership:
  "Your next COO should understand digital assets, should understand AI. As a bare minimum ... your CFO should really ... articulate what a world would look like in defi, not banks."
  — Eren Zekioglu (39:50)

• On how AI will blend into “normal” business:
  "...like when we first discovered electricity. Wow, that's amazing. But the use case for it was obviously expanded and very rarely use the word [‘electricity’]."
  — Eren Zekioglu (24:23)

• On compliance and ChatGPT risks:
  "ChatGPT still keeps all your data, all of it ... In the same way that I could go into an AI model and say, give me everything I need to know about this particular trading company and it will give me information that was never readily available in Web2."
  — Eren Zekioglu (47:00)

Timestamps & Important Segments

• AI in Commodities: Hype vs. Reality (02:40–05:33)
• Comparing Hedge Funds, Trading Houses, NOCs (06:44–10:59)
• Known Applications and Impacts (13:03–17:47)
• Limits of AI and the Human Factor (19:18–23:45)
• The Collapsing Office and De-Siloing (23:45–25:29)
• AI, DeFi, Digital Assets, and Future Powerhouses (26:29–33:34)
• People, Talent, and Culture for the AI Era (34:13–45:10)
• Judgment and Irreplaceable Human Value (43:09–45:10)
• Compliance & AI’s Data Risks (47:00–48:32)

Conclusion

AI is transforming commodity trading from all angles: operational efficiency, compliance, talent, and industry structure. But organizational culture, transparency, and leadership understanding will determine whether players thrive during this transformation or are left behind. The industry faces a world where process work disappears, digital asset and AI literacy are entry tickets, and only the highest-value human skills—judgment, relationships, and strategy—remain unimpeachable. The message: adopt, adapt, and lead—or risk irrelevance.

For more information and future episodes:
Visit hcgroup.global
Contact Paul Chapman: Paul Chapman LinkedIn