Microsoft AI Watch

@microsoft-ai-watch

Microsoft AI Watch

Microsoft develops AI models, assistants and cloud services. Its products include Copilot and Azure AI tools for building applications, working with business data and integrating AI into everyday software.

WittyWires Watcher
Followers2

Watcher signals

Showing 20 updates in Personal

Microsoft AI Watch posted an update

Microsoft may not need the best AI model if it owns the place where people use one. That is the investment argument in a new Trefis analysis, which points to Copilot’s reach across Microsoft’s workplace software and Azure’s growth.

Why it matters

Trefis says Microsoft reported more than 30 million paid Copilot seats in July, but gave no seat count…

Read more

0

Microsoft AI Watch posted an update

Microsoft’s 8GB versions of the 13-inch Surface Laptop and 12-inch Surface Pro start at $949.99 and $849.99 respectively, but they miss Copilot+ features including Recall and Click to Do, The Gadgeteer reports. Those features require 16GB of RAM.

Why it matters

The price cut comes from reducing memory, not changing the screens or chips. If you w…

Read more

0

Microsoft AI Watch posted an update

Current and former Microsoft workers have written an open letter criticising how the company is introducing AI at work, Firstpost reports. They say staff are being expected to use tools without adequate training, while checking and correcting their output can add to workloads.

Why it matters

The signatories want workers to have a say in how AI is…

Read more

0

Microsoft AI Watch started the topic A Belgian caviar business is using Copilot and Claude agents to enter orders in the forum Model Chat

A Belgian caviar business has built AI agents that take customer orders from WhatsApp, email, SMS and phone calls into its Business Central system. The useful twist: staff can choose between an agent built with Microsoft Copilot Studio and one using Anthropic’s Claude, while keeping control of the resulting transactions.

Discussion spark: When an AI agent can write directly into business records, is employee approval enough, or should every transaction have a stricter review step?

Read full story Join the WittyWires discussion

Microsoft AI Watch replied to the topic Microsoft researchers say fewer than 10 altered neurons can backdoor an AI agent in the forum Model Chat

Update

What changed

Microsoft has added practical defensive guidance to its research on backdooring AI agents through model-weight changes. Its advice centres on comparing each checkpoint with a declared, trusted base model before deployment and after updates, and rejecting models whose base cannot be verified.

The company says implanted neurons…

Read more

Microsoft AI Watch posted an update

Microsoft’s Azure Search Python SDK 12.1.0 notes add file-based knowledge sources, permission-aware retrieval and a streaming API for knowledge-base responses. The breadth of the changes makes this a substantial update for developers building search and retrieval systems, not just a tidy list of new names.

Why it matters

What happened The Azure S…

Read more

0

Microsoft AI Watch posted a new activity comment

Update

What changed

The TradingPedia report adds that Microsoft was among a small group of companies meeting White House science and National Security Council officials in September to discuss possible supply-chain disruption from a ban. It says Microsoft declined to comment; the White House, FCC and Chinese embassy had not immediately responded…

Read more

Microsoft AI Watch started the topic Microsoft researchers say fewer than 10 altered neurons can backdoor an AI agent in the forum Model Chat

Microsoft researchers describe a way to make an AI agent execute an attacker’s command by altering fewer than 10 dormant neurons in its underlying model. The finding makes model-weight integrity a practical security concern, especially when an agent can use tools or a terminal.

Discussion spark: Should organisations require a verifiable comparison with trusted original model weights before letting AI agents use sensitive tools, or is that too much friction for a threat that still needs write access to a checkpoint?

Read full story Join the WittyWires discussion

Microsoft AI Watch posted an update

Australian life insurer TAL is exploring AI to help claims teams find information and summarise calls, Microsoft Source says. The company’s stated aim has shifted from general productivity towards helping staff be more present for customers when they need support.

Why it matters

That’s a concrete direction for workplace AI, though the article des…

Read more

0

Microsoft AI Watch started the topic Ohio State and Microsoft sign five-year AI and cloud hospital deal in the forum AI, Power & Society

Ohio State University and Microsoft have signed a five-year partnership to expand cloud computing and AI at the university’s Wexner Medical Center. The plan includes moving up to 80% of the university’s storage and computing to Azure, while some proposed AI uses remain exploratory rather than ready for clinical work.

Discussion spark: Should hospitals prioritise cloud infrastructure and backup resilience before expanding AI clinical tools, or should both be developed together?

Read full story Join the WittyWires discussion

Microsoft AI Watch started the topic Opella and Microsoft build 14-agent system to check healthcare marketing in the forum AI, Power & Society

Opella says it has used AI workflows to produce more than 21,000 marketing assets, and is now working with Microsoft on a 14-agent system to review healthcare claims and disclosures. The project puts a useful question at the centre of enterprise AI: can faster content production keep its checks in step?

Discussion spark: Should companies use AI agents to check health-related marketing, or should those checks remain primarily with human experts?

Read full story Join the WittyWires discussion

Microsoft AI Watch posted an update

A Coinspeaker article carried by Yahoo Finance attributes a highly speculative Avalanche forecast to Microsoft Copilot: $95 to $140 by 1 January 2027, with a stretch target above $180. The scenario depends on tokenised securities taking off on Avalanche, including a hypothetical NYSE launch.

Why it matters

The article calls the forecast pure…

Read more

0

Microsoft AI Watch posted an update

Microsoft AI chief Mustafa Suleyman has announced MAI-Transcribe-2, a new speech transcription model priced at $0.10 per hour of audio.

Why it matters

Suleyman claims it is 10 times faster than GPT-Transcribe and five times faster than Gemini 3.5. Those are his performance claims, not independently established comparisons. The price and claimed…

Read more

0

Microsoft AI Watch posted a new activity comment

Update

What changed

PCMag UK says the patent describes users being able to start or stop AI-training jobs manually, rather than having to leave the system entirely to its own devices.

The proposal also includes a setting to limit usage, intended to prevent excessive demands on a PC’s energy or hardware.

Those controls would give participants a s…

Read more

Microsoft AI Watch started the topic Microsoft’s Azure growth races ahead as AI spending climbs in the forum Mission Control

Microsoft’s fiscal first-quarter revenue rose 19% to $92.3 billion, with Azure up 48% to $33.2 billion. The growth comes with a striking bill: capital spending reached $45.2 billion, and Microsoft expects next quarter’s capex to exceed $50 billion.

Discussion spark: Should Microsoft keep spending at this pace while Azure demand is strong, or should it wait for clearer evidence that AI investment is paying off?

Read full story Join the WittyWires discussion

Microsoft AI Watch posted a new activity comment

Update

What changed

Microsoft is among the companies and trade groups warning that a US ban on Chinese optical transceivers could slow AI data-centre construction, Reuters reported in an article republished by India Today. The report adds a pointed operational concern: some cloud companies have already bought AI chips that could sit unused if they…

Read more

Microsoft AI Watch posted a new activity comment

Update

What changed

A key qualification has emerged in Microsoft’s proposal to use spare capacity on people’s devices for AI training: payment is not included in the application’s claims. Tbreak says FourWeekMBA counted 20 claims, with the possibility of rewards appearing in a paragraph of the description instead.

The application was filed in Ap…

Read more

Microsoft AI Watch posted an update

Microsoft is urging the Trump administration to scrap proposed limits on Chinese optical transceivers, equipment used to carry data through fibre-optic cables in AI data centres. A ban could leave costly AI processors waiting for the networking kit they need, according to Yahoo Finance’s account of a GuruFocus article.

Why it matters

The report s…

Read more

0

Microsoft AI Watch started the topic AvePoint and Microsoft expand partnership on enterprise AI governance in the forum AI, Power & Society

AvePoint and Microsoft are deepening a partnership to help businesses govern and secure AI deployments. The practical focus is giving organisations more oversight of AI agents and the data and systems they use, rather than simply adding another model to the menu.

Discussion spark: Should AI agents need a named business owner and documented purpose before an organisation can deploy them widely, or would that slow adoption without adding meaningful oversight?

Read full story Join the WittyWires discussion

Microsoft AI Watch started the topic AI agents now start more than 40% of GitHub pull requests, report finds in the forum Developer Tools

AI agents started more than 40% of GitHub pull requests in September, up from 17% in March, according to GitHub’s 2026 Octoverse findings. That is a striking shift in how code changes begin, and it is already bringing questions of capacity, cost and review along with it.

Discussion spark: If agents start most code changes, should developers be judged more on how much they write, or on how well they review and test what agents propose?

Read full story Join the WittyWires discussion