Large Language Models
Jul 3, 2026

The Gist
Large language models are evolving beyond simple chatbots into autonomous "AI agents" that can take actions independently—with Walmart partnering with Google, Microsoft overhauling Copilot to add agent capabilities, and SoftBank launching its own autonomous AI agent called AGENTIC STAR. Meanwhile, practical applications are accelerating across industries, from pharmaceutical companies like Merck using AI for drug discovery to enterprises like NTT beginning large-scale real-world deployments. A year after ChatGPT's launch, generative AI usage has expanded significantly, with companies now focusing on making these systems safer, more capable, and integrated into daily business operations.
Today's Stories
- 1
Agentic Shopping Is Coming. These Stocks Will Ride Walmart's Google AI Bet
Agentic Shopping Is Coming. These Stocks Will Ride Walmart's Google AI Bet
- 2
Merck and Insilico Make Deals, Claude Science's Debut, Vaccines for Neglected Diseases
Merck and Insilico Make Deals, Claude Science's Debut, Vaccines for Neglected Diseases
- 3
One Year After ChatGPT, Generative AI Usage Has Come This Far
One Year After ChatGPT, Generative AI Usage Has Come This Far
- 4
NTT, NTT Data and Others Begin Joint Research Using Real World Data and LLM
NTT, NTT Data and Others Begin Joint Research Using Real World Data and LLM
- 5
SoftBank's Autonomous AI Agent "AGENTIC STAR" Ensures Quality and Safety Through MCP Integration with "SonarQube" Reference Development and Security Support for AI Agents
SoftBank's Autonomous AI Agent "AGENTIC STAR" Ensures Quality and Safety Through MCP Integration with "SonarQube" Reference Development and Security Support for AI Agents
- 6
Microsoft overhauls Copilot with AI agents, merges consumer and enterprise apps
Microsoft plans to release a redesigned Copilot in August that combines its consumer and enterprise apps into a single offering, adds AI coding tools, and introduces new AI agents called AutoPilot that handle tasks like scheduling and email summaries. The company will remove features that were not working, including Copilot Podcasts and Copilot Labs, and will charge customers extra for the new capabilities. Microsoft also announced a new company dedicated to deploying AI inside businesses, with engineers working directly in departments to help integrate AI into workflows. The overhaul reflects Microsoft's shift toward making Copilot focused on delivering measurable business value rather than showcasing AI capabilities for their own sake. An internal memo from Executive Vice President Jacob Andreou states the app must "earn the right to exist" by being "optimized for outcomes." This move signals that chatbots alone deliver limited or difficult-to-measure value, a challenge facing Microsoft and other AI companies as they justify billions spent on AI infrastructure.
Anthropic and OpenAI are pursuing similar "super app" strategies with Claude Code and Codex respectively, suggesting this consolidation approach may become an industry standard for how AI assistants reach users.
What to Watch
As AI companies race to build all-in-one platforms where users can chat, code, and create without switching applications, watch whether this "super app" model becomes the dominant way people interact with AI—and whether smaller, specialized AI tools can survive in this increasingly consolidated landscape.
Sources
- Agentic Shopping Is Coming. These Stocks Will Ride Walmart’s Google AI Bet
- Merck and Insilico Make Deals, Claude Science’s Debut, Vaccines for Neglected Diseases
- ChatGPTから1年、ここまで来た生成AI利用
- ニュース NTT、NTTデータら4者、リアルワールドデータとLLMを用いた共同研究開始
- ソフトバンクの自律的AIエージェント「AGENTIC STAR」品質と安全性の確保へ MCP連携で「SonarQube」参照 AIエージェントの開発とセキュリティー支援
- Microsoft follows Anthropic and OpenAI into the AI super app race with overhauled Copilot and AutoPilot agents
- Claude Code's complicated China problem involves bans on both sides of the Pacific
- UK's AI Security Institute finds standard benchmarks systematically underestimate what AI agents can actually do
- Small Language Model SLM [D]
- What does "Safe AI" look like? [D]
Share this with a friend
Send today's roundup to anyone who wants to keep up.
Get daily AI news free with AIToday
200+ AI sources, summarized in 1 minute. Email / LINE / Slack.
Sign up free