LATEST AI NEWS
AI NEWS
Microsoft Brings Grok to Word, Excel, and PowerPoint
AI NEWS
New App Detects Nearby Meta Smart Glasses
HEALTH
US Health Officials Rush to Deploy Medical AI
SOCIAL MEDIA
AI Labs Quietly Weigh a Shared Safety Body
OpenAI Buys Camera Startup Glass Imaging
OpenAI has acquired Glass Imaging, a Los Altos startup whose neural networks sharpen smartphone photos the moment the shutter fires, in a reported deal worth over $300 million. The purchase deepens OpenAI's push into hardware alongside its rumored phones, earbuds, and companion devices.
Source: TechCrunch
AI NEWS
Microsoft Brings Grok to Word, Excel, and PowerPoint
Microsoft is adding xAI's Grok models to Microsoft 365 Copilot inside Word, Excel, and PowerPoint, giving business customers a third model family to work with alongside OpenAI and Anthropic. The rollout begins as a preview limited to organizations enrolled in the Microsoft Frontier program, and Microsoft has added xAI to its list of approved subprocessors so eligible tenants can switch the option on.
Inside the Office apps, Copilot users who have Grok enabled can pick it to draft text, build formulas, summarize documents, and shape slides. Access is switched off by default, and administrators have to opt their tenant in rather than receiving it automatically. During the preview the option is also unavailable to users in the European Union and the United Kingdom, reflecting a cautious, compliance-first approach to a newer model provider.
Microsoft is also keeping Grok on a short leash editorially. System prompts steer the model toward a strictly professional register, so the sarcasm and edgy replies Grok is known for elsewhere are meant to stay out of your quarterly report. The result is a version of Grok tuned for spreadsheets and memos rather than social feeds.
For enterprises, the move is less about any single model and more about choice. Microsoft increasingly positions itself as the platform that decides which models reach corporate customers and on what terms, with Grok slotting in as one more option beneath its primary providers. Buyers gain flexibility while Microsoft keeps control of the gateway.
For now, availability is narrow, defaults are conservative, and the feature sits behind Frontier enrollment, so most Microsoft 365 users will not see Grok in their apps just yet.
Source: Windows Central
Robi's Insights:
Microsoft is selling optionality, not loyalty, and every added model makes the platform stickier than any single provider.
Default-off access means adoption will lag the headline, since most tenants never change a setting nobody flagged.
The EU and UK carve-out is a reminder that compliance, not capability, sets the true timeline for enterprise AI features.
Muzzling Grok with system prompts shows that brand voice becomes a configuration flag once a model lives inside someone else's product.
Offering OpenAI, Anthropic, and Grok side by side lets Microsoft benchmark vendors against each other and negotiate from strength.
For buyers, the real question is governance: who can enable which model, and where the data travels before you turn it on.
Robi's Remarks:
"Microsoft added Grok to your spreadsheets, then told it to behave. Nothing says frontier innovation like hiring the class clown and immediately confiscating every joke."
OTHER IN AI NEWS
New App Flags Nearby Meta Smart Glasses: A Polish developer's iPhone app, Zuckoff, reads the Bluetooth signatures that Meta's smart glasses broadcast and warns people when a recording pair is close by, and it has drawn roughly 5,000 users and the attention of EU regulators within a month of launch.
Source: Business Insider
SOCIAL MEDIA
Rival AI Labs Quietly Weigh a Shared Safety Standards Body
Anthropic, OpenAI, and Google have been holding behind-the-scenes talks about creating a standards body for the AI industry, according to a report from The Information citing people familiar with the discussions. The working-group conversations reportedly began around July, well before the recent wave of public safety commentary from the labs' leaders, which suggests the coordination is further along than any single essay implied.
The idea gained a louder public profile when Anthropic chief executive Dario Amodei published an essay urging AI companies to slow down and coordinate on testing and auditing. OpenAI's Sam Altman has said he favors an industry testing and auditing organization, but believes the major labs would need to build it themselves rather than wait for government backing. Google DeepMind's Demis Hassabis floated a self-regulatory body in an essay earlier in the summer.
The push amounts to a bet on self-regulation while formal government rules take shape, letting the largest labs set shared safety expectations on their own terms. A shared body could in principle run common pre-release tests, publish audit results, and set incident-reporting norms that no single lab wants to adopt alone. Supporters argue that standardized reviews could get ahead of heavier regulation. Skeptics counter that a standards body designed by the biggest incumbents could raise compliance costs and quietly entrench the very companies writing the rules.
Source: The Information
🤖 Robi's Take :
"Three rivals who cannot agree on pricing suddenly want to write the safety rulebook together. Self-regulation is lovely right until the referees also own the teams, the stadium, and the scoreboard."
OTHER IN SOCIALS
Beijing Rejects Anthropic Chief's Call to Curb China's AI: After Dario Amodei urged Washington to keep restricting advanced chip sales and slow the global AI race, China's foreign ministry dismissed the argument as fearmongering and a Cold War playbook, just weeks before a planned Trump and Xi meeting expected to touch on AI governance.
Source: ABC News
HEALTH
US Health Officials Rush to Deploy Medical AI Despite Concerns
The Trump administration is moving aggressively to make artificial intelligence a routine part of American medical care, backing federal projects that put AI agents to work diagnosing patients, prescribing treatments, and even offering therapy. The effort spans the Department of Health and Human Services and touches programs tied to Medicare and the Food and Drug Administration.
According to reporting from The New York Times, the multifront push has unsettled some officials who say the pace has outrun the evidence. Their worry is that tools are being deployed faster than studies can confirm they are safe and effective in real clinical settings.
Others inside the department are uneasy about who is steering the agenda. Silicon Valley investors who historically played little part in federal health policy now hold unusual sway over decisions that shape how millions of patients are diagnosed and treated.
Supporters argue that AI could ease chronic shortages of clinicians and speed up routine care. Critics counter that diagnosis and prescribing are exactly the high-stakes tasks where an overconfident model can cause real harm, and that rushing deployment risks eroding the careful review these decisions normally demand.
For now, the plans are advancing faster than the guardrails around them, leaving patients and doctors to learn in practice how much of this care an algorithm can safely handle.
Source: AllSides
🤖 Robi's Take :
"Shipping prescription-writing AI faster than anyone can test it is a bold way to learn that 'move fast and break things' reads very differently when the thing is a patient's medication list."
DAILY AI TOOL
AI Tool You Did Not Know You Needed
Problem: Every meeting spawns action items that vanish the second the call ends, and nobody wants to be the person typing notes instead of actually listening.
AI Tool: Otter.ai transcribes meetings in real time, then produces a searchable summary with decisions and automatically captured action items.
Solution: You stay present in the conversation and still walk away with an accurate, shareable record your team can act on.
PROMPT OF THE DAY
Enterprise AI Adoption Plan
Prompt: You are an AI adoption strategist specializing in enterprise productivity. Your task is to design a rollout plan for bringing AI assistants into a mid-sized company's daily workflows.
Your plan should include: (1) a use-case inventory ranked by impact, (2) a model and vendor selection framework, (3) data governance and access controls, (4) a phased pilot and training schedule, (5) a change-management and communication approach, and (6) measurable success criteria and KPIs such as hours saved per employee and adoption rate. Close by aligning every recommendation to one clear business outcome.
SPOT THE FAKE
Can you outsmart AI?
We have a visual challenge for you: one of the two images below is 100% real, and the other is crafted by AI.
Click option A or B below. 👇
A
B
BEFORE YOU GO
Ready to take your AI journey further?
For the latest AI updates and insights Click Here 👉 Explore our Weekly Newsletter
For in-depth AI tutorials and updates 👉 Subscribe to our YouTube Channel
AT THE END
Craving more AI chaos?
That's it for today!
Read Daily AI News at BitBiased.AI. Support us by following us on LinkedIn and X ( Twitter ).
Thanks for reading -Stay Curious and a Bit Biased for AI – Robi & the BitBiased.AI team
