Reeling from the software meltdown? Here's how to AI-proof your portfolio

The wipe-out in software stocks got me thinking: is there such a thing as an AI-proof portfolio?

We’re already seeing that AI disruption is real, but uneven. Where value is digital, replicable, and intermediary-based, risk is rising. Where value is physical, regulated, or tied to human necessity, AI has the potential to strengthen incumbents rather than displace them. 

More specifically, if capital is flowing out of stocks perceived to be vulnerable to disruption, where is that capital likely to flow instead? 

By “AI-proof”, I don’t mean a portfolio that is untouched by AI. Rather, I mean one where companies may be enhanced by AI, but ultimately deliver something fundamental to human existence. 

The starting point, then, is to figure out what humans will always do, no matter how advanced the technology around us becomes. Ultimately, the meatbag that is a human being requires certain inputs and creates certain outputs that can never be fully substituted - no matter how polite the interface is when it says, “Hello Dave”.

Here goes.

1. We will always eat
2. We will always need energy
3. We will always live in physical spaces (unless we work out how to upload consciousness to the cloud - in which case, my superannuation concerns will change materially)
4. We will always move ourselves and goods
5. We will always get sick and age (we may live longer, but Father Time remains undefeated for roughly 200,000 years and counting)
6. We will always generate waste (unfortunate, but unavoidable)
7. It could be argued that the compliant among us will always manage risk and follow the rules
8. We will always seek entertainment, identity, and status (even in a world where AI and robots are the great leveller, it’s easy to imagine coveting the T-1000 over the T-100 for the same reason people upgrade iPhones: “It’s got a better camera”)

I’ve probably missed a few, and I’m keen for readers to add to the list in the comments, but it’s a solid starting point for stock identification.

Important note: Whilst there are many stocks mentioned below, they should not be considered recommendations. They simply fall into the categories above. And importantly, “AI-proof” (or even AI-resistant) does not mean valuation-proof. Further analysis is required to determine whether any of these are good investments.

1. Humans will always eat and drink

Even with the rise of GLP-1s and the subsequent re-emergence of the heroin-chic aesthetic, food production is getting harder, not easier. Global population and consumption continue to rise.

Eating requires:

• Arable land
• Fertilisers and crop protection
• Heavy machinery
• Storage, logistics, and processing
• Retail distribution

Whilst AI may help with yield optimisation - drought resistance, disease mitigation, and the like - it still requires nutrients, equipment, and physical throughput.

And for the record, if the future involves Soylent Green, I’m out.

Stocks across the food stack

Retail & distribution

• Woolworths Group (ASX: WOW)
• Coles Group (ASX: COL)

Farming (land, inputs and storage)

• Australian Agricultural Company (ASX: AAC)
  
• Graincrop (ASX: GNC)
• Ridley Corporation (ASX: RIC)
• Nufarm (ASX: NUF)
• Nutrien (TSE: NTR)
• Bayer AG (ETR: BAYN)

Machinery & equipment

• Deere & Company (NYSE: DE)
• CNH Industrial (NYSE: CNH)

Processing & staples

• Nestlé (SWX: NESN)
• Bunge (NYSE: BG)
• PepsiCo (NASDAQ: PEP)

Fast food retailers

• McDonald’s (NYSE: MCD)
• Yum! Brands (NYSE: YUM)
• Collins Foods (ASX: CKF)

Bottom line: AI doesn’t replace fertiliser, diesel engines, or land scarcity.

2. Humans will always need energy

Aside from the structural demand required to keep lifting an increasingly growing population above the poverty line, for which energy safety is the primary determinant, let’s not forget that AI massively increases electricity demand - the machines don’t run on vibes. So there are multiple tailwinds for this category. 

What energy actually requires

• Fuel sources
• Transport and storage
• Transmission networks
• Maintenance of ageing infrastructure

Stocks across the energy chain

Energy networks

• APA Group (ASX: APA)
• NextEra Energy (NYSE: NEE)
• National Grid (LON: NG)

Traditional energy supply

• Woodside Energy (ASX: WDS)
• Santos (ASX: STO)
• Saudi Arabian Oil Co (TADAWUL: 2222)

Uranium & baseload

• Cameco (NYSE: CCJ)
• Paladin Energy (ASX: PDN)

Bottom line: The world - and AI - runs on energy. It cannot be substituted.

3. Humans will always live in physical spaces

What housing and cities require

• Construction materials
• Engineering and maintenance
• Transport corridors
• Water and waste systems

Stocks linked to physical civilisation

Infrastructure & roads

• Transurban Group (ASX: TCL)
• Atlas Arteria (ASX: ALX)

Construction & materials

• James Hardie (ASX: JHX)
• Heidelberg Materials (ETR: HEI)

Engineering & services

• Monadelphous (ASX: MND)
• Downer EDI (ASX: DOW)

Bottom line: AI doesn’t pour concrete or maintain bridges.

4. Humans will always move themselves and goods

What movement requires

• Vehicles
• Fuel
• Fixed transport corridors
• Ports and terminals

Stocks exposed to physical movement

Rail & freight

• Aurizon (ASX: AZJ)
• Union Pacific (NYSE: UNP)

Air & travel

• Qantas Group (ASX: QAN)
• Auckland International Airport (ASX: AIA)

Shipping

• Maersk (CPH: MAERSK-B)
• COSCO Shipping (SHA: 601919)

Bottom line: Routing will improve with AI, but asset owners still earn the return.

5. Humans will always get sick and age

What healthcare requires

• Medicines and devices
• Hospitals and clinics
• Regulation and trust

Stocks tied to biology, not software

ASX healthcare

• CSL (ASX: CSL)
• ResMed (ASX: RMD)
• Cochlear (ASX: COH)
• Sonic Healthcare (ASX: SHL)

Global healthcare

• Novo Nordisk (CPH: NOVO-B)
• Johnson & Johnson (NYSE: JNJ)

Bottom line: AI supports medicine. It doesn’t replace biology or regulation.

6. Humans will always generate waste

What waste requires

• Collection networks
• Permits and landfills
• Long-term environmental management

Stocks covering waste management

• Reece (ASX: REH) (think about it, toilets and plumbing)
• Cleanaway Waste Management (ASX: CWY)
• Waste Management (NYSE: WM)
• Republic Services (NYSE: RSG)

Bottom line: You cannot automate away physical waste.

7. Humans will always manage risk and comply (mostly) with rules

What compliance requires

• Capital
• Regulation
• Trust and distribution

Insurance & compliance stocks

ASX

• Insurance Australia Group (ASX: IAG)
• Suncorp Group (ASX: SUN)
• QBE Insurance (ASX: QBE)

Global

• Chubb (NYSE: CB)
• Zurich Insurance Group (SWX: ZURN)
• Berkshire Hathaway (via Geico, etc.) (NYSE: BRK.B)

Bottom line: AI helps pricing; regulation preserves the moat.

8. Humans will always seek entertainment, identity, and status

What still matters

• Live experiences
• Brands
• Scarcity

Stocks with non-digital moats

ASX

• Aristocrat Leisure (ASX: ALL)
• Viva Leisure (ASX: VVA)

Global

• Live Nation Entertainment (NYSE: LYV)
• LVMH (EPA: MC)
• Ferrari (NYSE: RACE)
• Formula One Group (NASDAQ: FWONK)

Bottom line: AI increases the supply of content. History suggests that when supply explodes, scarcity and brand matter more, not less.

Well, there you have it. A handful of deeply human behaviours that are unlikely to disappear, and a set of businesses that sit at critical points along those value chains. If nothing else, this thought experiment should challenge how you think about risk and resilience in a portfolio. 

Whether AI ultimately proves revolutionary or incremental, markets are already aggressively repricing parts of the economy. In that environment, it pays to understand where disruption ends - and where it doesn’t even really begin.

Over to you

How are you thinking about portfolio resilience given the onset of AI disruption? Which fundamental human endeavours did I miss and, most importantly, which stocks do you like that are "AI-proof"?