Microsoft Shifts Dynamics 365 and Power Platform Plans to AI at Work Roadmap

IBM has released its Granite 4.2 family of open reasoning language models in 3B, 8B, and 30B parameter sizes. These models feature native reasoning, a thinking/non-thinking switch, and a low-effort mode. They are pre-trained from scratch and fine-tuned for specific tasks, making them suitable for applications like chatbots and virtual assistants.

Meanwhile, Microsoft is shifting its Dynamics 365 and Power Platform product plans to the AI at Work roadmap, improving visibility for customers and partners. This move aims to provide a more transparent and predictable approach to product planning.

Cigna is leveraging AI to cut healthcare costs and improve patient outcomes. The company estimates that its AI and predictive analytics tools can save $275 billion in annual revenue. However, a report warns that the increasing use of debt to fund AI investments has reached $445 billion, posing risks to the economy.

Scale AI and Microsoft are not the only ones making significant investments in AI. Private equity firm Southfield Capital predicts that AI will eliminate traditional corporate departments like finance and marketing within five years. The firm is investing in custom AI solutions that bypass conventional departmental structures.

In Southeast Asia, generative AI is becoming popular in online shopping, but human recommendations still play a bigger role in purchasing decisions. However, AI agents pose serious security risks in business environments, often going unnoticed by security teams.

Key Takeaways

• IBM releases Granite 4.2 AI models with native reasoning and a thinking/non-thinking switch. • Microsoft shifts Dynamics 365 and Power Platform plans to the AI at Work roadmap. • Cigna's AI and predictive analytics tools can save an estimated $275 billion in annual revenue. • AI-related debt has reached $445 billion, raising concerns about economic risks. • Private equity firm Southfield Capital invests in custom AI solutions to eliminate traditional corporate departments. • Generative AI is becoming popular in Southeast Asia's online shopping journey. • AI agents pose serious security risks in business environments. • China cracks down on AI companions, citing concerns about emotional dependence and loneliness.

IBM Unveils Granite 4.2 AI Models with Native Reasoning

IBM has released Granite 4.2, a family of open reasoning language models in 3B, 8B, and 30B parameter sizes. These models are built around explicit reasoning and can emit a chain of thought before answering. They also feature a thinking / non-thinking switch and a low-effort mode. The models are pre-trained from scratch and have been fine-tuned for specific tasks.

IBM's Granite 4.2 Models for Local LLMs

IBM has made Granite 4.2 models available for download, including 3B, 8B, and 30B variants. These models offer a 128,000-token context window and some support agentic reinforcement learning. They can be used for applications like chatbots, virtual assistants, and language translation systems.

Generative AI in Southeast Asia's Online Shopping

A report by Cube Asia finds that generative AI is becoming popular in Southeast Asia's online shopping journey, but human recommendations still play a bigger role in purchasing decisions. Consumers in the region rely heavily on creators, friends, family, and other human voices when deciding what to buy.

Invisible AI Agents Pose Security Risks

AI agents can pose serious security risks in business environments, often going unnoticed by security teams. These agents can read inboxes, file tickets, write code, and move records between applications. A report by Reco found that small and mid-size companies carry 414 unsanctioned AI tools per 1,000 employees.

Doncaster Patient Frustrated with AI Receptionist

A woman recovering from a stroke was frustrated with an AI reception service at her GP surgery, which could not understand her. The patient had to repeat her name and details multiple times before the call dropped off. The surgery has since decommissioned the AI system.

Private Equity Firm Bets on AI to Eliminate Departments

Private equity firm Southfield Capital predicts that AI will eliminate traditional corporate departments like finance and marketing within five years. The firm is investing in custom AI solutions that bypass conventional departmental structures.

Cigna Uses AI to Cut Healthcare Costs

Cigna's chief data, digital and AI officer, Katya Andresen, is using AI to cut healthcare costs and improve patient outcomes. The company has announced that its AI and predictive analytics tools can save an estimated $275 billion in annual revenue.

The AI Debt Binge Threatens the Economy

The increasing use of debt to fund AI investments is raising concerns about the potential risks to the economy. Large-scale AI models require immense computing power, leading companies to take on significant debt. Analysts estimate that AI-related debt has already reached $445 billion.