London Tech Week 2027 (7 to 11 June): the UK AI adoption deals to watch
London Tech Week 2027 runs 7 to 11 June, with expo days 7 to 9 June at Olympia. The stage plan, the 2026 deals that set the bar and the AI adoption signals to track.
Topic
Enterprise AI strategy — maturity models, build-vs-buy, architecture (RAG, fine-tuning, agents), governance, and the discipline that turns AI into measurable impact.
London Tech Week 2027 runs 7 to 11 June, with expo days 7 to 9 June at Olympia. The stage plan, the 2026 deals that set the bar and the AI adoption signals to track.
What NRF has published for Retail's Big Show 2027: the agentic commerce sessions, named speakers, pass prices and deadlines, what is still unannounced, and what to prepare.
A date-by-date 2027 retirement calendar for 15 Claude models on the Claude API, Amazon Bedrock, Microsoft Foundry and Google Cloud, with the earliest date per model and a migration plan.
MPMS pays on sales above a moving baseline, not total sales. The TS1 window opened on 15 October 2026. What year one, ending 31 March 2027, means for brands and EMS firms.
Notice periods, minimum lifetimes, legacy access and auto-upgrade rules for OpenAI, Anthropic, Gemini API, Agent Platform, Bedrock and Foundry, with a dated 2027 calendar.
What Nasscom has confirmed for its 35th NTLF, what its FY26 numbers say about growth, hiring and AI revenue, and the questions CTOs should take into 2027 vendor talks.
Google scaled ads in AI Mode to India on 28 October 2026. What Search, Shopping and Performance Max advertisers can control, what they cannot see, and how to plan 2027 budgets.
Nova Lake and Medusa are announced but unspecified. What Intel and AMD have confirmed, what ships today, and a buy-now-or-wait plan tied to the 12 October 2027 Windows deadline.
Instinct's Terms of Service, revised 20 August 2026, take a perpetual, irrevocable licence over screen captures and keystrokes. Workspace data is carved out, so the same email can fall on both sides.
Mistral released Shieldstral 1.0 as Apache 2.0 open weights on 4 August 2026. On Mistral's own benchmark tables, GPT-OSS-Safeguard-20B leads 9 of 18 text rows to Shieldstral's 4.
Vercel's 21 August 2026 changelog says every v0 connection mints a short-lived token with no long-lived secret to rotate. Its own Connect docs describe two connector types that store exactly that.
Version 1.0.0 of the anthropic Python package landed on 20 August 2026 and moved the HTTP layer to httpx2. The upgrade installs, the type checks pass, and the mocks quietly stop matching.
Anthropic shipped local session transcript endpoints on 11 August 2026, exposing Cowork and Claude Code conversations to reviewers. Five common setups return nothing, and the export loop skips messages.
Anthropic brought memory stores to self-hosted Managed Agents sandboxes on 19 August 2026. It is a download-and-sync copy, not a mount, and the limits are specific enough to change how you run workers.
Unity Catalog ABAC can now mask columns and filter rows for requests an agent makes on a user's behalf. Databricks names three access paths the control misses, starting with personal access tokens.
Databricks set an end-of-life date of 30 September 2026 for the Supervisor API (Beta). Neither the AWS nor the Azure August 2026 release notes list the deprecation, and Beta is on by default on Premium.
AWS says deny by default is available wherever Amazon Quick is. Its own region table lists 26 regions and marks 8 for agentic features. Mumbai, Ohio and Singapore are not among them.
Anthropic added allowed_domains and blocked_domains to Claude Managed Agents web tools on 20 August 2026. The Console organization-level list does not reach agent sessions and each list caps at 64 entries.
OpenAI now lets one API key target us., eu. or in.api.openai.com per request. At launch only the US, EU and UAE ran inference in region. India, Japan and Australia have since added it for gpt-6-luna only.
A2A reached 150 supporting organisations and a stable 1.0 specification in 2026, while MCP shipped a stateless protocol core on 28 July 2026. Most enterprise agent systems need only one of them. Here are five tests that
Google Cloud's DORA team models a 39% first-year return on AI-assisted development, then tells you not to treat it as a formula. What it assumes, the productivity dip it hides, and the cost line most rollouts miss.
Open weights closed the intelligence gap from 13 points to 6, yet enterprise share fell to 11%. New August 2026 data shows identical weights scoring 22% or 37% on tool calling depending only on the endpoint.
Machine identities outnumber humans 109 to 1 and 12% of organisations have given AI agents privileged access to core systems. Okta's Permiso deal shows where agent identity governance is heading.
OpenAI's internal red-teamer beat human red-teamers 84% to 13% and broke a live vending-machine agent. GPT-5.6 Sol now fails 0.05% of those attacks. Here is why a hardened model is not a defence plan, and the seven