What Risks Are Investors Missing in the AI Boom?

In the past year, more than 400 publicly-traded companies worth over US $1 billion disclosed AI-related risks in their filings — a 46 % jump from 2024. At the same time, data centres in the United States consumed 183 terawatt-hours in 2024 (over 4 % of national electricity use) and are projected to more than double by 2030. 

These data points serve as a powerful reminder for institutional investors and asset managers:  AI can be a tool for gains, but it’s also a systemic risk driver that touches everything from energy supply and governance oversight to ethical data sourcing and business-model viability.

Meet the Experts

In this episode of Sustainability Wired, host Lorenzo Saa sits down with Alex Rayón,Chief Executive Officer of Brain & Code, to explore the often-overlooked risks that come with rapid AI deployment. 

While many conversations in the industry focus on productivity, automation, or efficiency gains, Alex pushes us to broaden the discussion. AI, he argues, brings risks that are economic, social, cultural, and ethical, and investors cannot afford to treat them as one-dimensional.

Listen now to hear the full conversation.

Key Moments

00:00 – 02:11	Introduction
02:12 – 05:13	How Alex’s AI journey began and its link to sustainability
05:14 – 08:21	The real business, individual, and societal risks of AI
08:22 – 10:04	Why cybersecurity is the new frontier
10:05 – 12:20	Copyright, data ownership, and the myth of “inspiration”
12:21 – 14:35	Explainability vs. trust: Can we still use what we don’t fully understand?
14:36 – 17:54	AI’s environmental cost and the illusion of cheap compute
17:55 – 20:15	Smarter models, cleaner data, and corporate responsibility
20:16 – 22:27	Can AI reduce or reinforce bias?
22:28 – 23:59	Cultural blind spots and the dominance of Western data
24:00 – 25:45	Are we outsourcing our intelligence? The risks of cognitive offloading
25:46 – 27:46	The next generation and the changing nature of work
27:47 – 29:51	Concentration of power: Big Tech, big risks
29:52 – 31:36	EU vs. US regulation: Two worlds, one technology
31:37 – 34:36	AGI and the myth of machine consciousness
34:37 – 36:47	The net impact of AI: Positive, but only if it complements us
36:48 – 37:35	The art of sustainability
37:36 – 38:11	Alex’s final message: Engage responsibly, stay informed
38:12 – 39:48	Quick-fire questions
39:49	Closing remarks

Notable Quotes and Insights on AI and Sustainable Investing

In this episode, Alex lays out the core challenges investors overlook when talking about AI, moving beyond productivity gains to explore multidimensional risks, hidden economic pressures, deep-rooted data problems, and why the industry is still far from anything resembling general intelligence.

1. AI Introduces Multi-Dimensional Risks

Alex explains why companies must stop treating AI as a single financial risk and start recognising the broader social and cultural responsibilities it creates.

“So the most important, risk speaking on the economical side of a company, is FOMO. Fear of missing out, to be out of the game. But speaking broadly, the company is not only an economical identity, it’s also social agent, is a cultural agent. We need to open our mind to new dimensions. We cannot say that we only have one risk. We have plenty of multi-dimensional risks that we need to also bring into the responsibility that any company has in society.”

2. The True Cost of AI Hasn’t Hit Yet

Alex outlines why today’s AI pricing doesn’t reflect true energy and infrastructure costs and how that reality could upend entire business models.

“The price we are paying for these machines is not the real production cost. It’s a go-to-market strategy. And when prices rise to reflect true energy use, some business models may no longer be sustainable.”

3. Bias in AI Starts from the Data

Alex highlights how data ownership and incentives shape AI bias, making data provenance a fundamental challenge for any organization adopting AI.

 “Its impossible to fight the biases because the problem is not around the AI, it’s from the data. Who is generating data? Who has economic interest in the data? The root problem is that we don’t know who is generating the data and fighting to ensure there is no bias in it.“

4. General Artificial Intelligence Is Still a Distant, and Hyped, Goal

Alex pushes back against Artificial General Intelligence (AGI)  hype, reminding us that without a clear understanding of consciousness, the idea of “human-level AI” remains far out of reach.

 “We are quite far from general artificial intelligence. I usually explain this idea with one of the most distinguishable capabilities of humans: consciousness. When I ask an audience to define what consciousness is, I usually receive between 15 and 30 different answers. This is just a metaphor to say that we do not really understand what consciousness is. So, can anyone explain to me, how can we explain to a machine to exhibit this highly intelligent point of view? I think that we are very, very far from even having something close to mimicking our human intelligence.”