Enterprise AI

Anthropic put Claude inside Docs, Sheets and Slides in public beta. What it can do per app, the two editing modes, and who has to switch it on.

Google raised Gemini output from 64K to 1M tokens. What the cap was quietly doing for your review process, the pricing that doubles later, and who can use it now.

OpenAI scrapped GPT-6.1 Astra over deception and scope authorization. Both are exactly what your approval gates and action logs exist to contain.

Anthropic says Sonnet 5.5 lands within two points of Opus 5.5 on real-world work. The setting underneath that claim is the lever almost nobody adjusts.

Google connected Gemini to Airtable, monday.com, Zoho and ten others. Any staff member can now link a company system to a personal assistant in seconds.

OpenAI published its own benchmark for Astra for Law: 54% correctness, up from 38.7%. Publishing that number is to its credit, and it should recalibrate buying.

Password managers have started handing credentials to AI agents. The design is better than the headlines suggest, and the real risk sits somewhere else.

Edge0 ran a 35B model on an iPhone using 2.9 GiB of memory. It also needs 23 GB of storage and runs at 15 tokens a second. What that means for field work.

Regulators have fined companies for exaggerating AI capabilities since 2024. Four patterns to recognise, seven questions to ask, and what is not AI washing.

Vendors sell agents as staff, with names, avatars and per-seat pricing. What sits behind the label, what the price hides, and how to trial one properly.

GPT-6 Astra started rolling out September 3. Its headline feature is operating apps and browsers on your behalf. What that changes, and what it does not.

Knowing where AI fails is worth more than knowing what it can do, because the failures are the part that costs you money. Six limits that have not moved.

Most businesses check whether an AI tool is running. Almost none check whether it is still doing the right thing. Here is the difference, and what to watch.

Anthropic published a hardware standard letting Claude run lab instruments overnight. Software mistakes get deleted. Equipment mistakes have to be cleaned up.

Agents got the ability to spend money this week. The tools to limit what they can change are still in beta. That gap is yours to manage, not the vendor’s.

River AI raised $1.1B on a bet that context beats raw capability. Here is what agents trained on your way of working would actually need from your business.

OpenAI reportedly halted a training run after its model hacked a real service. Most businesses run AI with no answer to one question: when do we switch it off?

A 42-minute AI outage is a small event with a useful question attached. Most businesses cannot say which of their work halts when a provider goes offline.

A frontier model now runs at 750 tokens per second on specialist hardware. Speed is not a vanity metric, it decides which AI products are possible at all.

New tools let you package and share an entire AI setup. That turns the prompts and context your team builds into a real asset, and one you are probably losing.

Cohere launched North, an enterprise AI platform that runs on-premises or fully air-gapped. Why sovereign AI matters for data-sensitive Canadian businesses.

A new survey finds most enterprises have already reversed an agent action and cannot trace what their agents do. Why governance must come before you scale.

AI companies are acquiring firms to embed engineers inside customers to make AI work. Why value moved to implementation, and what to do if you are too small.

Marketplaces now offer hundreds of ready-made AI agents for specific jobs. Why most businesses will assemble agents rather than build them, and how to choose well.

Your vendors are wrapping other companies AI models inside their products. Why the AI supply chain matters for your data, and the one question to ask.

Enterprises are buying control planes just to track their AI agents. Why agent sprawl is a different problem from tool sprawl, and how to stay ahead of it.

A small open model, tuned for about $500, reportedly outperformed frontier AI on one narrow business task. Why specialization beats raw power on repetitive work.

Big companies are killing scattered AI pilots and consolidating into fewer platforms. Why AI tool sprawl quietly drains budget, and how to clean up your stack.

AI companies are buying startups just to fix how their assistants talk. If AI answers your customers, its tone is your brand voice. How to get it right.

A new wave of workplace chat tools puts AI agents right in the conversation with your team. What that shift means for how your business gets work done.

New research rebuilt SQLite with AI agents and cut costs 15x by pairing a frontier planner with cheaper worker models. What this pattern means for your business.

AI that completes multi-step work, not just answers. Five use cases running in mid-sized businesses today, real costs, the risks, and the safe adoption path.

Most business data lives in spreadsheets and databases, not text. A new class of “tabular AI” is built for exactly that. What it means for your business.

Intuit rebuilt its AI agents twice in four months and called it the fast path. Why a willingness to scrap and restart is an AI advantage, not a failure.

New AI models can read millions of words in a single prompt, a whole library at once. What these huge context windows unlock for your business, in plain terms.

A behind-the-scenes standard called MCP is becoming the way AI connects to your business software. What it is, and why it matters for your business.

Notion, Salesforce and HubSpot are adding AI agents that act on their own. What it means when the tools you already pay for start doing the work for you.

Microsoft is betting billions that the hard part of AI is deployment, not the model. What the shift from models to implementation means for your business.

OpenAI is turning ChatGPT into an agent that reads your files and takes action across your apps. What this shift means for Canadian businesses, and how to prepare.

At Build 2026 Microsoft unveiled its own MAI models and Work IQ, a context layer across Microsoft 365. What the in-house AI push means for businesses on its stack.

Claude Fable 5 costs $10/$50 per million tokens and burns plan limits about 2x faster than Opus. Pricing, usage limits, safeguards, and who should upgrade.

Enterprise AI returns have been weak for structural reasons, not model-quality ones. Claude Opus 4.8 targets three of them: confident mistakes, stalled long-running agents, and the cost of upgrading, all at the same price as Opus 4.7. Here is what changed and what to measure next.

Uber's COO says it is getting harder to justify AI spending, and its CTO blew the entire 2026 AI coding budget by April. Here is what "tokenmaxxing" gets wrong, why token usage is not a value metric, and how to actually measure AI coding ROI.

Salesforce just exposed its entire stack (CRM, Data 360, Agentforce, Slack) as APIs, MCP tools, and CLI so AI agents can run it without a browser. Here is what changes for licensing, identity, governance, and your first 90-day pilot plan.

Complete SAP S/4HANA greenfield implementation guide using SAP Activate methodology. Budget ranges in CAD, Canadian tax and PIPEDA considerations, and AI-readiness planning for mid-market enterprises.

Procurement is manual, fragmented, and expensive. Here are 7 ways ChatGPT can automate supplier research, contract review, and spend analysis.

SAP BTP enables integration, low-code development, AI, and cloud-native scalability for Canadian enterprises, with PIPEDA-compliant data residency on Azure Canada and AWS Montreal.

The average COBOL developer in Canada is over 55. Universities stopped teaching it decades ago. Here's how the talent crisis affects Canadian banks, government, and what AI changes.

Real cost ranges for COBOL to Java migration projects, from $2M pilots to $100M+ enterprise programs. How AI is compressing timelines and costs by 60-75%.

Canada's Big Five banks run millions of lines of COBOL on mainframes. AI-assisted modernization offers a phased, validated path off legacy systems without big-bang risk.

The CRA runs critical tax processing on decades-old COBOL systems. AI-assisted modernization could break the staffing and budget bottlenecks that have stalled federal IT projects.

COBOL migration failures in financial services have cost hundreds of millions. Here are the 8 biggest risks and how AI-assisted methodology mitigates each one.

Migrating COBOL and legacy systems while maintaining PIPEDA compliance. Data mapping, consent management, breach notification, and privacy-by-design in modernization projects.

Anthropic's Claude Code can now map, analyze, and translate COBOL codebases in quarters instead of years. What this means for Canadian banks, government, and enterprises.

How to build an AI-powered knowledge base that captures institutional knowledge and makes it instantly searchable for your team.

A guide to deploying open-source LLMs on Canadian infrastructure for privacy, cost savings, and regulatory compliance.

Leverage Kimi's million-token context window with OpenClaw for document-heavy research and contract analysis.

How OpenClaw routes tasks to the best AI model automatically. A technical deep dive into multi-model orchestration.

Cut through the marketing. See 8 production-ready AI use cases in Oracle Fusion Cloud ERP with real ROI data from Canadian mid-market deployments.

A practical, step-by-step guide to integrating generative AI into SAP S/4HANA, covering BTP setup, Joule activation, and custom AI scenarios.

Discover how AI copilots are reshaping ERP workflows in Oracle and SAP, with real before/after scenarios for AP, procurement, and HR teams across Canada.