Top Companies' AI Moves

Today's Stories

1. 1

   Consumer Reports investigated how Uber and Lyft use AI algorithms to determine pricing, raising questions about fairness and transparency in ride-hailing fares.

   Consumer Reports launched an investigation into the AI systems that Uber and Lyft use to set prices for rides. The investigation examined how these algorithms work and what factors they consider when calculating fares. Riders have little visibility into how their fares are determined, and understanding whether these AI pricing systems are fair and transparent has become a consumer protection concern. The investigation sheds light on practices that directly affect what passengers pay.

   The results of Consumer Reports' investigation may prompt questions about whether ride-hailing companies should disclose more information about how their pricing algorithms function and whether regulators should impose greater transparency requirements on these systems.

2. 2

   Caterpillar is positioned to capitalize on surging power demand driven by AI data centers, as major cloud providers seek reliable backup power solutions.

   Caterpillar has begun supplying diesel generator sets to hyperscalers (large cloud providers) seeking backup power for AI data centers. The company is also starting to serve ads on its platforms and has introduced new products including modular power units and carbon capture solutions. AI data centers require continuous, dependable electricity, and diesel generators serve as critical backup when grid power fails. Caterpillar's entry into this market puts it at the center of infrastructure spending driven by hyperscalers' rapid expansion of AI capabilities.

   Caterpillar's ability to scale production of power solutions while competing with other equipment suppliers. The company's modular power units and carbon capture offerings may also determine its relevance as data centers face pressure to reduce emissions.

3. 3

   ServiceNow's workflow software may become more valuable, not less, as companies deploy AI agents—because those systems still need centralized management and governance.

   ServiceNow reported that AI-related products such as Now Assist continued to gain traction, with Now Assist's net new annual contract value more than doubling year over year in the fourth quarter of 2025, and delivered 21% revenue growth in 2025. Many investors feared AI agents would replace workflow software like ServiceNow's. In reality, AI agents can identify problems and generate solutions, but organizations still need workflows to approve spending, notify suppliers, update systems, and track every action. ServiceNow positions itself as the platform to manage both human employees and AI systems from a single place—what management calls an 'AI Control Tower.'

   The company faces real competition from new AI-native start-ups and existing enterprise software rivals, and some investors believe future AI agents could automate far more of companies' workflows than expected. However, ServiceNow starts from a position of strength because its software is deeply embedded in many of the world's largest organizations.

4. 4

   ISG is holding an event to help companies measure the business value they're getting from their AI investments.

   ISG (a business advisory firm) is organizing an event focused on how enterprises can turn their AI investments into measurable business outcomes. The event will explore practical approaches to quantifying AI's impact on operations and strategy. Many companies are spending heavily on AI but struggle to prove actual business returns. This event addresses a real gap—helping leadership teams and buyers understand whether their AI spending is delivering concrete results, not just technological capability.

   The event brings together enterprise leaders and decision-makers to share how they are measuring and demonstrating AI's value to their organizations.

5. 5

   DeepSeek releases a new reasoning model that becomes the #2 open-weights reasoning model, offering capabilities approaching larger competitors at a lower computational cost.

   DeepSeek released DeepSeek-R1, which becomes the #2 open-weights reasoning model. The model uses 32T–33T tokens during training and requires 83.9 GiB of VRAM, making it smaller and less resource-intensive than competing models while delivering strong performance on reasoning benchmarks. Open-weights reasoning models (AI systems that can work through complex problems step-by-step, with publicly available code) have been dominated by proprietary options. DeepSeek's release at a lower computational threshold may broaden access to advanced AI capabilities for organizations without massive infrastructure budgets.

   The model is available for free download and use, removing cost barriers to entry. Its performance relative to larger closed-source reasoning models will determine whether it reshapes how businesses evaluate their AI infrastructure spending.

6. 6

   DeepSeek's open-weights reasoning model has reached #2 ranking, shaking up the competitive landscape as smaller AI labs prove they can challenge the largest players.

   DeepSeek released an open-weights reasoning model that achieved #2 ranking among open-weights reasoning models. The model uses 32T–33T tokens and operates on 27% of FLOPs compared with DeepSeek-V3.2. Open-weights models (AI systems whose underlying code is publicly available) are lower-cost alternatives to proprietary systems. A strong #2 ranking signals that well-resourced independent labs can now compete directly with the largest AI companies on reasoning tasks, which is a capability that was previously seen as concentrated among a few well-funded players.

   The model's efficiency metric — achieving competitive performance on roughly one-quarter of the computational work of its predecessor — suggests that the cost barrier to building advanced AI systems is lowering, which could reshape investment and vendor lock-in dynamics in the sector.