CIBC Tech & Innovation Market Update – August 2024

In late July, computer systems all over the globe were brought to a halt, with many users experiencing the “blue screen of death”. Businesses across a variety of sectors including healthcare, government, technology, and transportation were left unable to perform day-to-day operations after a crash of CrowdStrike’s software impacted Microsoft Windows operating systems. As of Monday, it was reported that up to 8.5 million Windows devices were affected by the outage; however, with a significant number of Fortune 500 companies running CrowdStrike software on Windows systems, the impact was magnified.

Meta recently announced that the company’s latest AI model will be unavailable in Europe given the regulatory environment. The tech giant became the subject of controversy after announcing plans to train AI models on public social media data, resulting in a series of complaints filed to data protection authorities across 11 European countries. Despite offering users in Europe the ability to opt out in an attempt to comply with the EU’s General Data Protection Regulation (GDPR), Meta was eventually ordered to pause training on EU user data. As a result, the company elected to withhold access to the model in the EU rather than continue adjusting its product in hopes of achieving compliance. Apple recently took action against European regulators in a similar fashion: the company announced last month that some iPhone features will be unavailable in the EU due to regulatory concerns. European regulators have been heavily cracking down on Big Tech this year, introducing the Digital Markets Act (DMA) which specifically targets large online platforms with more than 45 million monthly active users. The responding actions of both tech companies now highlight the growing tensions between US-based tech giants and European regulators.

It’s no surprise that generative AI workloads require a lot of energy. We wrote about the energy consumption of large language models and data centers last year, and in that report we credited the hyperscalers for setting and maintaining carbon-free energy targets. However, news from Google this week indicates the tide may be turning when it comes to how AI is impacting those ESG targets. Bloomberg reported that Google has ended its mass purchase of carbon offsets, and ceased claiming that Google’s operations are carbon neutral, something the company has claimed for nearly two decades (since 2007). Instead, Google has said it now aims to achieve net-zero emissions by 2030. While the tech giant stated a commitment to achieving its sustainability goals, the strategy pivot could still be viewed as an indication that other carbon targets are at risk. Google separately referred to its net-zero target as an “extremely ambitious goal”.