A 60-minute walkthrough of Anthropic's new desktop agent. We'll cover what Cowork is and how it differs from chat, how to set up and run your first agent from scratch, real agent examples you can adapt, and practical agent-design patterns to use from day one.

Whether you're new to Cowork or already pushing past one-off setups into real agent workflows, you'll leave with something usable.

Presented by Josh Sullivan, COO at Kiingo AI

Thursday, May 14, 2026
10:00 AM PT / 12:00 PM CT / 1:00 PM ET
Save your spot here →

Note: requires a paid Claude plan.

• Microsoft and OpenAI ended their seven-year exclusive cloud deal on April 27. OpenAI can now distribute its models on any cloud, Microsoft retains a non-exclusive license to OpenAI IP through 2032, and the revenue share runs through 2030 with a cap. GPT-5.5, GPT-5.4, Codex, and Bedrock Managed Agents went live on AWS Bedrock the next day. (Microsoft, AWS)

• Anthropic had a major week. It's reportedly in talks for a $50 billion round at a $900 billion valuation, which would surpass OpenAI ($852B) as the most valuable AI startup in the world — up from $380B in February, with annualized revenue now at $30B. It also launched a new enterprise AI services company alongside Blackstone, Hellman & Friedman, and Goldman Sachs, targeting mid-sized organizations that lack in-house resources to deploy AI — community banks, manufacturers, and regional health systems. Separately, Claude Security (powered by Opus 4.7) entered public beta, scanning codebases for vulnerabilities with integrations into CrowdStrike, Palo Alto Networks, SentinelOne, Microsoft Security, and Wiz. (CNBC, Bloomberg, Anthropic, TechCrunch)

• Microsoft Agent 365 is now generally available at $15 per user per month, or bundled in the new Microsoft 365 E7 SKU at $99/user/month. It gives IT one place to discover, govern, and secure AI agents across Microsoft and third-party tools, including detection of "shadow" agents employees deploy without approval. Runtime Protection in public preview, plugs into Entra ID and Defender. (Microsoft Security, VentureBeat)

• Meta raised 2026 AI capex guidance to $125–145 billion and acquired humanoid robotics startup Assured Robot Intelligence (ARI) — the latest signal that hyperscalers are extending their AI bets into physical systems, not just software. (Fortune, TechCrunch)

• OpenAI hit four milestones at once. It locked in 10 gigawatts of US AI compute capacity — roughly enough to power 7–8 million homes — three years ahead of its 2029 target, with 2GW from Amazon plus sites in Texas, New Mexico, Wisconsin, and Michigan. ChatGPT for Clinicians launched free for verified U.S. physicians, NPs, PAs, psychologists, and pharmacists with pre-built clinical workflows. ChatGPT Workspace Agents flips from free preview to credit-based pricing on May 6 for Business, Enterprise, Edu, and Teams. And GPT-5.5-Cyber began rolling out through the Trusted Access for Cyber program to government, infrastructure operators, and financial firms. (Bloomberg, OpenAI: Clinicians, OpenAI: Workspace Agents, OpenAI: Cyber)

• Back-office agents went mainstream this week. Salesforce launched Agentforce Operations for process coordination, data verification, compliance reviews, and approval routing (vendor claims 50–70% cycle-time cuts and 80% fewer manual errors). IBM made "Bob," an AI development partner that handles planning through deployment, generally available; IBM's lead reference customer cut a 30-day Java upgrade to three days. (Salesforce, IBM Newsroom)

• The agent economy got real infrastructure this week. Cloudflare opened its platform to AI agents — they can create accounts, register domains, get API tokens, and deploy code, with Stripe co-designing the provisioning protocol. Stripe's new Link digital wallet lets autonomous agents make purchases with a built-in $100/month default spending cap per provider. And OpenAI's new Agentic Commerce Protocol (open-source, codeveloped with Stripe) powers Instant Checkout in ChatGPT, letting users buy from Etsy and soon Shopify merchants directly inside the chat. (Cloudflare, Stripe, OpenAI)

• Independent benchmarks of GPT-5.5 show a clear lead on end-to-end coding tasks (82.7% on a standard agentic-coding test, ahead of Claude Opus 4.7 and Gemini 3.1 Pro), though accuracy varies meaningfully by use case. The practical takeaway: pick the model for the job. Coding-heavy work where GPT-5.5 leads is a different decision than research and fact-checking work, where verification layers still matter regardless of which model you use. (Artificial Analysis)

Three findings worth knowing this week.

• IBM surveyed 2,000 CEOs and found 76% now have a Chief AI Officer, up from 26% in 2025. 83% of CEOs say AI success depends more on people's adoption than the technology, and organizations that redesigned five core business areas are 4x more likely to deliver on business objectives. (IBM Institute for Business Value, May 2026)

• A Federal Reserve note found 18% of U.S. firms had adopted AI by year-end 2025, but employment-weighted, 78% of the labor force works at AI-adopting firms. Around 41% of workers report using generative AI at work. Firm-level adoption grew 68% annually. (Federal Reserve Board, April 2026)

• A Stanford Digital Economy Lab study of 51 successful enterprise AI deployments found 95% of AI transformation failures trace back to organizational factors, not technology. Workforce unpreparedness, missing governance, and absent executive ownership account for nearly all of it; tech underperformance explained fewer than 5%. (Stanford Digital Economy Lab, March 2026)

[Pick one initiative your team is about to deploy AI on, or has tried and stalled with. Tell me what it is, who owns it, what data it touches, what success looks like, and what's blocking it right now if you know.]

Walk through five organizational areas and tell me whether each is ready to support this initiative or will quietly break it. Cover: (1) Technology: does the data exist, is it accessible, is the integration path clear? (2) Finance: is the budget realistic for what this actually requires, and how is ROI measured? (3) HR: who is being asked to change how they work, and who's accountable for adoption? (4) Operations: are we redesigning workflows, or bolting AI onto existing process? (5) Cross-functional ownership: who breaks the tie when these teams disagree? For each area, tell me what's solid, what's missing, and what would need to be true for this not to fail. If three or more areas are clearly underprepared, tell me to pause the launch and fix the foundation first.

Return it as: ready / partially ready / not ready for each of the five areas + the single biggest gap to close before deploying.

I've spent enough time on the AI side of work that it follows me out of the office. Friends and family describe a task at dinner, and I'm halfway through assessing the workflow before they finish the sentence. There's almost always a fix. A faster prompt, a different tool, a question they hadn't asked.

I used to do the same thing as a lawyer. Friend would describe a situation at dinner and I'd be running it through "is there a legal issue here?" before I could react to it as a person. The reflex doesn't go away. It just switches domains. Now I'm spotting workflow problems instead of liability ones, and the answer is almost always "yes, AI could fix that in ten minutes."

What still surprises me is how invisible this is to people who haven't been in it. The gap between "AI is everywhere" in the news and "I'm actually using it on the thing in front of me" is enormous, and most people I talk to are still on the wrong side of it. Not because they're slow on the uptake. Because nobody has shown them what's possible with tools they already pay for.

If you're still skeptical, sit down with someone who uses these tools well, hand them something on your plate that's been bugging you, and watch what happens. That's usually the moment. It's worth digging in.

Most AI initiatives don't stall on the technology. They stall on the work underneath: clean data, redesigned workflows, governance, accountable owners. That's where Kiingo helps. Training, strategy, implementation.

If you want to know whether your team is set up to deliver on the AI work in front of you, that's exactly the conversation we have. kiingo.com/contact