Microsoft Google Amazon Plan $520 Billion AI Investment as Energy Demands Grow

The artificial intelligence sector is experiencing unprecedented growth, with major technology companies like Microsoft, Google, Amazon, and Oracle planning to invest a staggering $520 billion in AI next year. This immense sum surpasses South Korea's entire proposed national budget for 2026, highlighting the rapid expansion and perceived potential within the AI industry. Such significant funding is expected to accelerate AI development and drive new innovations, while also raising questions about how national research efforts can compete with the financial power of these private tech giants. This rapid AI expansion also brings substantial infrastructure demands, particularly for energy. Electricity use is projected to increase by about 25% after years of stability, largely due to AI and new data centers. Meta Platforms Inc. is actively seeking US government permission to trade electricity to power its energy-intensive AI data centers, noting that power generation isn't keeping pace with demand. Microsoft is pursuing similar approvals, while Apple already has them. For instance, Meta's Louisiana data center alone will require at least three new gas-powered plants, underscoring the scale of energy needs. However, the proliferation of AI also presents significant risks and challenges. Major insurance companies, including AIG, Great American, and WR Berkley, are asking US regulators to exclude AI-related liabilities from business insurance policies. They cite the unpredictable and complex nature of AI models, fearing potential multibillion-dollar claims and widespread simultaneous problems if a common AI tool malfunctions, as seen with Google's AI Overview falsely accusing a solar company. One expert noted that AI model results are simply too unclear to understand fully. Beyond corporate liabilities, AI's impact extends to individual users and societal structures. Health insurance companies are increasingly using AI to process claims, contributing to a rise in denials, with approximately 73 million Americans on Affordable Care Act plans having in-network service claims rejected in 2023. In response, patients are now leveraging AI tools to challenge these denials. Furthermore, the Cardano network recently experienced a temporary split after an attacker used an AI-generated script to exploit a software bug, causing a 6% drop in its ADA cryptocurrency, though no user funds were lost. Globally, there's a push for responsible AI development. Germany and Korea are advocating for cooperation on human-centered AI, emphasizing respect for human dignity, civil liberties, and the rule of law. While Korea often highlights AI's economic opportunities, Germany focuses on dangers like misinformation. Similarly, Dr. Joshua Budhu from Memorial Sloan Kettering Cancer Center stresses the importance of designing AI for fair neurology care, warning that biased data or unfair use could worsen existing health disparities. He recommends involving diverse patients in AI development and ensuring clear rules for its use. Companies like Apple are prioritizing AI advancements, with the upcoming iOS 27 update set to focus on performance improvements and significant upgrades to its artificial intelligence capabilities, aiming to catch up in the AI field. Meanwhile, AI pioneer Llion Jones, who helped develop generative AI at Google, advises the UK to adopt a "brave" approach, focusing on unique AI research and fostering environments that grant researchers more freedom, rather than directly competing with US and Chinese tech giants. He also suggests that current AI chatbots need the ability to disagree with and correct users.

Key Takeaways

• Microsoft, Google, Amazon, and Oracle plan to invest $520 billion in AI next year, exceeding South Korea's entire 2026 national budget.
• Major insurers like AIG, Great American, and WR Berkley are seeking to exclude AI-related risks from business policies due to AI's unpredictability and potential for multibillion-dollar claims, citing incidents like Google's AI Overview.
• AI and new data centers are projected to increase US electricity demand by 25%, leading companies like Meta and Microsoft to seek government permission for electricity trading, a practice Apple already utilizes.
• Meta's Louisiana data center alone will require at least three new gas-powered plants to meet its energy demands for AI operations.
• Health insurance companies use AI for claims processing, contributing to approximately 73 million denied claims in 2023, prompting patients to use AI tools to appeal these decisions.
• The Cardano network experienced a temporary split due to an AI-generated script exploiting a software bug, causing a 6% drop in ADA value, though no user funds were lost.
• Germany and Korea are collaborating to develop human-centered AI, focusing on respecting human dignity and civil liberties, with Germany emphasizing user-centered design and Korea leveraging its data.
• Dr. Joshua Budhu advocates for designing AI with fairness in mind from the start to improve neurology care and prevent exacerbating health disparities, recommending diverse patient involvement.
• Apple's upcoming iOS 27 update will prioritize performance improvements and significant upgrades to its artificial intelligence capabilities to catch up in the AI field.
• AI pioneer Llion Jones advises the UK to pursue unique AI research and foster researcher freedom, rather than directly competing with US and Chinese tech giants, and suggests AI chatbots should be able to disagree and correct users.

Insurers limit AI coverage due to high risk

Major insurance companies like AIG, Great American, and WR Berkley are reducing their coverage for artificial intelligence risks. They are asking regulators for permission to remove liability for AI agents and chatbots from their insurance plans. Insurers worry about potential multibillion-dollar claims because AI systems are complex and hard to predict. This cautious approach comes as AI becomes more common in many industries.

Insurers say AI is too risky to cover

Major insurers like AIG, Great American, and WR Berkley are asking US regulators to remove AI-related risks from business insurance policies. They find AI models too unpredictable to understand fully. The industry worries about many claims happening at once if a common AI tool makes a mistake. For example, Google's AI Overview once falsely accused a solar company. Insurers can handle one big loss, but not thousands of simultaneous problems caused by AI.

Top insurers want to exclude AI risks

Major insurance companies, including AIG, Great American, and WR Berkley, are seeking approval from US regulators. They want to remove liabilities related to artificial intelligence from their business insurance policies. One expert in the field stated that the results from AI models are too unclear to understand.

Meta wants to trade electricity for AI centers

Meta Platforms Inc. is asking the US government for permission to trade electricity. This move aims to power its growing artificial intelligence data centers, which need a lot of energy. Meta's head of global energy, Urvi Parekh, stated that power generation is not increasing fast enough to meet demand. Other tech companies like Microsoft are also seeking similar approvals, and Apple already has them. Meta's Louisiana data center alone will require at least three new gas-powered plants. By trading electricity, Meta can manage power prices and encourage new energy projects, helping it reach its goal of net-zero emissions by 2030.

AI data centers boost electricity demand

Electricity use is expected to increase by about 25% after many years of staying the same. This big jump is happening because of artificial intelligence, new data centers, and more things using electricity. Experts like Ethan Zindler from BloombergNEF, former Energy Secretary Dan Brouillette, and Sunrun CEO Mary Powell discussed that the United States will likely need all types of energy to meet this growing demand.

Patients use AI to fight denied insurance claims

Health insurance companies are using artificial intelligence to handle claims, which has led to more denials. In 2023, around 73 million Americans with Affordable Care Act plans had their in-network service claims rejected. Very few of these people tried to appeal the decision. Now, patients are starting to use AI tools themselves to challenge these denied insurance claims. Indiana University law professor Jennifer Oliva explained how this new approach helps patients.

Germany and Korea push for human-centered AI

German Ambassador to Korea Georg Schmidt and AI ethics expert Kinga Schumacher are calling for Germany and Korea to work together on developing human-centered artificial intelligence. They believe this cooperation will ensure AI respects human dignity, civil liberties, and the rule of law. Schmidt noted that Korea often highlights AI's economic chances, while Germany focuses on its dangers like false information. Schumacher suggested Germany's way of designing things with users in mind could help Korea's fast AI growth. Both countries have strengths, with Korea having more data and Germany having more experience with AI uses. They also discussed how AI can help aging societies if designed with their needs in mind.

AI can improve fair neurology care

Dr. Joshua Budhu, a Neuro-Oncologist at Memorial Sloan Kettering Cancer Center, shared ideas on using artificial intelligence for fair neurology care. He explained that AI can help with earlier detection of diseases and provide tools that are easy to understand for everyone, especially in communities that do not get enough medical care. However, he warned that if AI is not designed carefully, it could make existing health differences worse due to biased data or unfair use. Dr. Budhu and his team suggest three main steps: involve diverse patients in AI development, teach neurology staff about AI, and ensure clear rules for how AI is used and checked for fairness. They stress that designing AI with fairness in mind from the start is very important.

Cardano network splits after AI-scripted attack

On Saturday, November 23, 2025, the Cardano network temporarily split into two separate chains. This happened when an attacker used an AI-generated script to create a "malformed transaction" that exploited a bug in the network's software. Newer nodes accepted this transaction while older ones rejected it, causing the split. Cardano co-founder Charles Hoskinson stated it was a targeted attack by a former testnet participant. Developers quickly released a fix, and network operators upgraded their systems. Although no user funds were lost, the incident caused ADA to drop over 6%. An X user named "Homer J." claimed responsibility, saying he acted alone and did not mean to cause harm.

Big Tech AI spending exceeds South Korea budget

Major technology companies like Microsoft, Google, Amazon, and Oracle plan to invest a huge $520 billion in artificial intelligence next year. This massive investment is more than South Korea's entire national budget proposed for 2026, which is also around $520 billion USD. This shows how quickly the AI industry is growing and how much potential these companies see in it. Such large funding will likely speed up AI development and lead to new discoveries. This trend also brings up questions about how national research efforts can compete with the financial power of big private tech companies.

iOS 27 to focus on AI and performance

Apple's upcoming iOS 27 update will reportedly focus on making the system better and faster, rather than adding many new features. The company plans to prioritize improvements in performance and significant upgrades to its artificial intelligence capabilities. According to Mark Gurman, catching up in the field of AI is a very important goal for Apple with this new release.

AI pioneer urges UK to be brave in tech

Llion Jones, an AI pioneer who helped develop generative AI at Google, is urging the United Kingdom to be "brave" in its approach to artificial intelligence. He advises against trying to compete directly with the large AI companies in the USA and China. Jones, now the Chief Technology Officer of Sakana AI, believes current AI chatbots need to be able to disagree and correct users. He suggests that Wales and Britain should focus on unique AI research and create environments that give researchers more freedom. Business Secretary Peter Kyle MP also highlighted AI's potential to transform the economy during an event in Cardiff. The Welsh government has an AI strategy and announced AI growth zones.

Sources

• Insurers retreat from AI cover as risk of multibillion-dollar claims mounts
• AI is too risky to insure, say people whose job is insuring risk
• AI is too risky to insure, say people whose job is insuring risk
• Meta Seeks US Approval to Trade Electricity for AI Data Centers
• AI Data Centers and the Need for More Electrons
• How patients are using AI to fight back against denied insurance claims
• Germany and Korea must build human-centered AI
• Joshua Budhu: Artificial Intelligence for Equitable Neurology Care
• Cardano (ADA) News: Network Splits in Two as Attacker Uses AI-Generated Script to Exploit Bug
• Big Tech AI Investment Surpasses South Korea's Budget
• iOS 27 will reportedly focus on performance improvements and AI upgrades
• AI pioneer Llion Jones calls for UK to 'be brave' in tech race