🗞️ THE WEEKEND RECAP

Top Stories You Might Have Missed

💰 OpenAI Raises $8.3B at $300B Valuation: Dragoneer led the round with $2.8B as investors rushed to join OpenAI’s cap table. Now at $13B in revenue and 700M weekly ChatGPT users, the company eyes $20B by year’s end and a true for-profit future.

🔍️ Apple Building AI Search Tool: Testing on-device ‘Answer Engine’ to rival Google and ChatGPT with faster, privacy-focused results baked into its ecosystem.

🚀 Google Launches Deep Think in Gemini App: AI Ultra subscribers get access to Deep Think, a powerful reasoning model based on Gemini 2.5 that scored medal-level at the 2025 IMO. It uses parallel thinking to tackle tough math, science, and coding problems step by step.

🥇 Anthropic Overtakes OpenAI in Enterprise AI: Anthropic now leads with 32% of enterprise usage, up from 12% in 2023, while OpenAI drops to 25%. Claude’s coding prowess helped drive the shift, with enterprises favoring closed models over open-source competitors.

📉 Dev Trust in AI Tools Falls Despite Surge in Use: 80% of developers now use AI, but trust has dropped to 29% as “almost right” suggestions spark hard-to-find bugs. Stack Overflow reports AI-caused issues are driving devs back to the site for help.

🛑 Anthropic Blocks OpenAI from Claude: Anthropic has quietly revoked OpenAI’s API access to its Claude models, intensifying rivalry as both chase dominance in enterprise AI tools.

✨ Cohere Debuts Lean Vision Model for Enterprises: Command A Vision runs on just two GPUs yet beats GPT-4.1 and Llama 4 Maverick on visual tasks like OCR and ChartQA, making it ideal for reading charts, diagrams, and PDFs in enterprise workflows.

🚨 OpenAI Pulls Chat Sharing Feature After Privacy Uproar: ChatGPT conversations appeared in Google search results, revealing sensitive user info. Despite being opt-in, backlash forced OpenAI to remove the feature within hours, highlighting fragile AI trust.

🍏Tim Cook Rallies Apple Around AI Push: In a company-wide call, Cook admitted Apple’s late to AI but said it “must win,” urging teams to own the space and seize the moment.

🚧 NVIDIA’s H20 China Sales Stalled by Commerce Backlog: Despite July approval to resume some AI chip sales to China, NVIDIA still awaits licenses as internal delays and national security concerns stall progress at the Commerce Department.

🛡️ Cursor AI Patched Critical Prompt Injection Flaw: A now-fixed CVE-2025-54135 bug let attackers run commands via poisoned Slack or GitHub content. Version 1.3 adds safer allowlists, scrapping faulty denylists to prevent stealthy RCE and SSH key theft.

Enjoying our newsletter? Forward it to a colleague—
it’s one of the best ways to support us.

🤥 DISINFORMATION

AI Played a Role in 80% of 2024 Elections, Mostly To Deceive, Not Inform

Artificial intelligence was used in 80% of national elections in 2024, according to a new global database from the International Panel on the Information Environment. While some AI tools were deployed for benign tasks—like translating speeches or aiding voter ID—most content was generated anonymously or by foreign actors, with 67% categorized as harmful. In cases like Bangladesh and India, deepfakes and synthetic endorsements blurred lines between fact and fiction.

The impact ranged from confusion to chaos: Romania annulled and re-ran its presidential election due to AI-driven interference. Though the panel refrained from judging outcomes, its data confirms a changing political information environment—one where truth and trust are increasingly up for grabs. → Read the full article here.

🧓 CAREGIVING

AI Companions Offer Hope, and Raise Red Flags, for Dementia Care

Startups like NewDays and CloudMind are rolling out AI companions designed to support people with dementia, aiming to reduce loneliness and promote cognitive engagement through personalized conversations. Tools like “Sunny” and “BrightPath” mimic friendly chat, guide users through memory exercises, and provide caregivers with mood summaries and behavioral insights, functions that could help lighten caregiver load and offer emotional support.

But experts warn these systems may overpromise and underdeliver. There’s little hard data showing AI conversations match the clinical benefits of human-led therapy, and concerns about privacy, isolation, and chatbot hallucinations loom large. For now, AI companions are a promising supplement, but not a substitute, for human care. → Read the full article here.

⚛️ QUANTUM BREAKTHROUGH

Quantum AI Cracks Semiconductor Design, Outperforms Classical Models

Australian researchers have used quantum machine learning (QML) to model a notoriously difficult part of chip design—Ohmic contact resistance—marking the first time quantum AI has been successfully applied to semiconductor fabrication. Their new algorithm, QKAR, encoded experimental data into quantum states to detect patterns traditional models miss, outperforming seven leading classical methods.

This breakthrough could streamline chip manufacturing, especially in data-scarce scenarios, and help build faster, more efficient electronics. As quantum hardware matures, techniques like QKAR may reshape the future of chipmaking from the atomic level up. → Read the full paper here.

📽️ VIDEO

This Might Be OpenAI's New Open-Source Model...

Horizon Alpha, a fast, likely OpenAI-backed open-source model, excels at coding, image analysis, and creativity, but sometimes struggles with logic and honesty.

🧰 TOOLBOX

Smarter Support, Safer Identities, and Reliable AI Performance

🗒️ FEEDBACK

Help Us Get Better

🤠 THE DAILY BYTE

Meet Abi, the New Star at the Care Home

That’s a Wrap!

❤️ Love Forward Future? Spread the word & earn rewards! Share your unique referral link with friends and colleagues to unlock exclusive Forward Future perks! 👉 Get your link here.

📢 Want to advertise with Forward Future? Reach 550K+ AI enthusiasts, tech leaders, and decision-makers. Let’s talk—just reply to this email.

🛰️ Want more Forward Future? Follow us on X for quick updates, subscribe to our YouTube for deep dives, or add us to your RSS Feed for seamless reading.

Thanks for reading today’s newsletter—see you next time!

Matthew Berman & The Forward Future Team

🧑‍🚀 🧑‍🚀 🧑‍🚀 🧑‍🚀