Jobs, Power Grids, Politics

What happened
Goldman Sachs warned that accelerating AI adoption could push U.S. unemployment higher, estimating the technology accounted for 5,000 to 10,000 net job losses per month in the most exposed industries and 7% of planned layoffs in January. Companies across sectors are trimming staff to fund AI initiatives: Allianz plans to cut up to 1,800 travel-insurance jobs, Amazon has slashed 16,000 corporate roles, Autodesk will shed 7% of workers, Dow is cutting 13% and Nike, Pinterest, SEB and Telstra have announced similar restructurings.

Why it matters
Investors and executives worry AI could disrupt established business models. Costly automation projects are driving companies to shift capital and labour toward AI systems, creating near-term job losses and uncertainty for employees.

What's next
Expect more restructuring announcements as firms race to integrate generative models into workflows. Policymakers may face pressure to address displacement via retraining and social safety nets.

What happened
A Reuters column argued that market panic over AI-driven unemployment and economic collapse is overblown. Viral posts like the 7,000‑word “2028 Global Intelligence Crisis” and a 5,000‑word piece by Otherside AI’s CEO stoked fears of millions of job losses and deflationary collapse, sending the U.S. software sector down 20% in a month. The column notes that these scenarios lack hard data and that valuations now reflect more realistic expectations.

Why it matters
Investor psychology can swing between euphoria and doom. Alarmist narratives have outsized impact on markets, even when not grounded in evidence. Balanced analyses show AI is simultaneously aiding and replacing workers.

What's next
As more empirical research emerges, investors may reassess AI’s impact. Look for sober analyses and data-driven forecasts to temper extreme narratives.

What happened
U.S. stock indexes rallied, with the Nasdaq up 1.26%, as concerns about AI disruption receded and optimism over the technology’s potential returned. Nvidia reported quarterly revenue of $68.13 billion, beating estimates, helping the S&P Software & Services index jump 2.9. Investors are grappling with existential risks but noted that AI may enable workers rather than displace them, according to Richmond Fed President Tom Barkin.

Why it matters
Markets remain sensitive to AI news. Positive earnings and reassurance from policymakers can quickly reverse sentiment, underscoring the speculative nature of AI-related stocks.

What's next
Upcoming results from software makers like Salesforce, Intuit and Snowflake will test whether the rally holds. Watch for continued volatility as investors digest mixed signals.

What happened
Big Tech plans to spend more than $600 billion on AI in 2026, but analysts warn U.S. power grids cannot keep pace. AI data centres consume over a gigawatt each—enough to power hundreds of thousands of homes—and more than 46 facilities plan to build their own power plants to meet demand. President Trump told companies to provide for their own power in his State of the Union, citing concerns that AI could raise electricity prices. Grid operators like PJM warn of potential shortages by 2027 and are requiring large users to develop generation or accept curtailment.

Why it matters
AI’s expansion is constrained by physical infrastructure. Sky‑high energy demands could slow model training and deployment, raise carbon footprints and trigger regulatory scrutiny.

What's next
Expect tech firms to invest in on-site generation and energy‑efficiency innovations. Policymakers may accelerate grid upgrades and craft rules to allocate costs fairly between data centres and consumers.

What happened
The White House invited Microsoft, Amazon, Anthropic and Meta to a March 4 meeting to formalise a “Rate Payer Protection Pledge” aimed at shielding consumers from AI‑driven electricity price hikes. The pledge builds on commitments like Microsoft’s investment in new generation and follows President Trump’s directive that tech companies must build their own power plants. White House spokeswoman Taylor Rogers said companies will sign the pledge alongside the president, while Anthropic pledged to cover 100% of any power price increases caused by its data centres.

Why it matters
Rising energy demand from AI data centres is becoming a political issue. The pledge seeks to defuse voter anger by ensuring companies absorb higher costs rather than passing them to households.

What's next
Watch the March 4 meeting for details on cost‑sharing mechanisms and whether major tech firms commit to transparent reporting of their energy use.

What happened
Chinese AI lab DeepSeek denied Nvidia and AMD early access to its forthcoming V4 model, granting domestic suppliers like Huawei a head start. Sources say the model was trained on Nvidia’s restricted Blackwell chips in mainland China, potentially violating U.S. export controls. DeepSeek’s models have been downloaded over 75 million times, and Chinese downloads on Hugging Face now surpass those from any other country.

Why it matters
The move underscores geopolitical tensions in AI. By disadvantaged U.S. hardware suppliers, China aims to build a self‑sufficient AI stack while testing the limits of export controls.

What's next
Expect U.S. regulators to investigate how DeepSeek accessed prohibited chips and whether further restrictions are needed. Chinese labs may accelerate model releases to cement domestic dominance.

What happened
The U.S. Defense Department asked contractors including Lockheed to assess their reliance on Anthropic ahead of a deadline for the AI firm to relax safeguards that prohibit its technology from being used for autonomous targeting and surveillance. Defense Secretary Pete Hegseth warned that if Anthropic fails to comply, it could be labelled a supply‑chain risk or forced to change its policies under the Defense Production Act. Contractors were told to prepare for potential action after the Friday deadline.

Why it matters
This standoff pits AI ethics against national‑security demands. A forced policy change could set a precedent for government intervention in model usage restrictions.

What's next
Anthropic has until 5 p.m. ET Friday to respond. The outcome could influence contract decisions and spark broader debates about AI governance in defence.

What happened
OpenAI hired Ruoming Pang, a high‑profile researcher who oversaw AI infrastructure for Meta’s Superintelligence Labs and previously led Apple’s AI models team. Pang reportedly left Meta after aggressive recruitment by OpenAI and had joined Meta only seven months earlier from Apple. He reportedly commanded a compensation package worth more than $200 million.

Why it matters
The hire illustrates the fierce talent war among AI leaders. Top engineers who can build scalable agentic infrastructure are commanding massive pay packages.

What's next
Expect more high‑profile moves as companies race to secure scarce AI expertise. Regulators may scrutinise employment contracts as poaching intensifies.

AI continues to dominate corporate strategy, labor markets and regulatory agendas. February 25 marked a turning point as investors balanced hysteria with realism, governments focused on power infrastructure and ethics, and competition intensified across borders and boardrooms.