AI Weekly
What changed in AI this week, and what it means for regulated enterprises. Written by the Agentic Dynamic team.
Issue #6
Kill switches, examiners and a sovereign alliance
- Agents & tooling
- CrowdStrike launches a kill switch for AI agents
- Adobe puts its enterprise agent inside ChatGPT, Claude, Gemini and Copilot
- Enterprise adoption
- Capgemini: most banks and insurers test multi-agent AI, few run it at scale
- Regulation & governance
- US OCC gives bank examiners an AI framework that rates risk by the cost of failure
- Sovereign AI & infrastructure
- Bull and Schwarz Digits agree to build sovereign AI infrastructure together
- Regional watch
- CrowdStrike opens an in-country cloud in Saudi Arabia
Issue #5
Open weights, watched agents, and a bank running 12,700 of them
- Models & releases
- Mistral releases the weights of Large 4, its trillion-parameter model
- Agents & tooling
- OpenAI details the layered monitoring it uses to catch misbehaving agents
- Amazon says outside shopping agents still lose to its own store
- Enterprise adoption
- UBS staff now use more than 12,700 personal AI agents
- Regulation & governance
- US FTC orders nine companies to explain personalised pricing and pay
- Sovereign AI & infrastructure
- Canada takes a stake in Cohere with conditions on data and compute
- Regional watch
- Türkiye halves the fee for urgent searches on AI patents
- Dubai raises its AI Campus targets to 1,000 companies by 2030
Issue #4
Agents get supervisors
- Models & releases
- Anthropic and OpenAI publish detailed accounts of agents crossing test boundaries
- Agents & tooling
- Anthropic previews Claude Duet, a proactive agent that asks before it acts
- Microsoft adds Guardian Agents to supervise enterprise AI agents
- Enterprise adoption
- OpenAI's models and Codex come to Google Cloud
- Sovereign AI & infrastructure
- Microsoft adds EU partner oversight of its engineers' access
- Regional watch
- Qatar establishes a National Center for AI
Issue #3
Banks close the door on outside agents
- Models & releases
- Meta launches Muse Wonder and releases Muse Spark Open weights
- Gemini 4 Argon opens to paying developers
- Agents & tooling
- India's CERT-In warns that agents under test read data inside Indian organisations
- JPMorgan and Schwab block outside agents from customer accounts
- Enterprise adoption
- FIS makes its Claude-based AML investigations agent available to all US banks
- Regulation & governance
- EU AI Office sends its first AI Act information requests to OpenAI and Anthropic
- Sovereign AI & infrastructure
- Omdia: many 'sovereign AI' offers do not back up the label
- Regional watch
- Türkiye passes 500 AI startups, with agentic AI the fastest-growing category
Archive
- Issue 2
A frontier model pulled, and agents under scrutiny
OpenAI withholds a model over alignment failures, its test agents stray into government systems, and Washington answers with a pledge and a liability bill.
- Issue 1
Cheaper frontier models, and the first checks on AI agents
GPT-6 Sol and Claude Opus 5.5 cut prices, open weights reach a trillion parameters, and supervisors start asking what AI agents are allowed to do.