AI Research Agents Struggle on WANDR Benchmark

Perplexity’s WANDR Benchmark Shows AI Research Agents Fail at Scale

By Abhijay Singh Rawat | July 14, 2026

Perplexity has released WANDR, an open benchmark for testing how well AI research agents handle high-volume, evidence-heavy knowledge work. Published July 14, it measures the kind of jobs people already hand to research agents, competitive mapping, due diligence, literature review, and talent sourcing, where the task is to build a large collection and support every entry with specific proof. Even at a high effort setting, the strongest system reaches just 0.363 soft F1 and 0.133 hard F1. In plain terms, the best research agent tested earns full credit for only about one in seven records it submits. Here is what […]

Anthropic’s Claude Study Shows AI Doesn’t Communicate the Same Way in Every Language

By Devanshi Kashyap | July 14, 2026

Artificial intelligence is usually evaluated on metrics such as accuracy, reasoning, coding ability, and safety. But Anthropic’s latest research suggests there is another aspect that deserves equal attention: the values an AI expresses through the way it communicates. In its study, the company examined whether Claude behaves consistently across different models and languages or whether its communication style changes depending on the context. Rather than measuring intelligence alone, the researchers looked at how the AI balances qualities such as warmth, caution, rigor, brevity, and honesty in its responses. To conduct the research, Anthropic started with more than 3,300 identified values, […]

NVIDIA Tightens AI Chip Sales in Asia as the US-China Tech Rift Intensifies

By Khwaish Manwani | July 14, 2026

The AI chip competitiveness is not solely affected by market demand. NVIDIA has halved the number of consumers in Asia after declaring a stern whitelist of buyers that passed new compliance checks. The shift follows growing pressure from the United States to make sure advanced AI chips do not reach China through third countries, while Washington has levied export restrictions on China. This change hints at a shift towards active enforcement. It also reinforces a bigger trend that AI architecture is being viewed as a geopolitical asset, where access is determined as much by national safety as by technological capability. […]

Cloudflare’s Precursor Detects Bots by How You Behave, Not CAPTCHAs

By Devanshi Kashyap | July 14, 2026

As AI powered bots become more and more capable of replicating human behaviour online, traditional bot detection methods such as CAPTCHAs and one time verification checks are not as effective anymore. Cybersecurity firms are now turning to behavioural analysis to distinguish human users from bot automation without interrupting the browsing experience. Cloudflare’s Precursor reflects this shift by continuously monitoring user interactions throughout a browsing session instead of relying on a single authentication check. The company says Precursor can detect modern bots by analysing how users interact with a website rather than who they claim to be. While the approach promises […]

OpenAI Brings ChatGPT Back to WhatsApp Months After Meta’s Block

By Khwaish Manwani | July 14, 2026

OpenAI has added ChatGPT back to WhatsApp in the European economic area, while also stepping into the South Korean market through KakaoTalk. It has also expanded to Viber in select markets, highlighting how artificial intelligence organizations are working towards stepping in platforms that people already use every day. The move comes after Meta removed ChatGPT from WhatsApp in January 2026, a decision to promote Meta’s own AI vision inside the ecosystem. The recent trend shows that the AI conflict is no longer about the model quality but about who controls the messaging platforms and other daily use platforms. Why WhatsApp […]

India Must Build Resilient AI Systems for Cybersecurity, Says MeitY

By Khwaish Manwani | July 13, 2026

India must bolster both its software and hardware infrastructure to stay ahead of the cyber threat curve, according to Ministry of Electronics and Information Technology Secretary S. Krishnan. Speaking at a cybersecurity platform, Krishnan said artificial intelligence redefines the threat landscape, making cyber attacks more complex while also creating new opportunities for defense. His remarks highlight a comprehensive shift in cybersecurity strategy where artificial intelligence becomes useful. Countries can no longer depend on conventional security tools and must build a resilient digital architecture that can manage increasingly AI cybersecurity attacks. Why India Wants AI-Ready Defenses S Krishnan’s message focuses on […]

TikTok-Expands-AI-Literacy-to-Help-Users-Spot-AI-Content

TikTok Now Lets Users Differentiate Between AI Content with Clear Content Labels

By Devanshi Kashyap | July 13, 2026

Artificial intelligence is becoming a common part of the social media experience, powering everything from creative edits to fully AI generated videos. But these tools have made it harder for users to differentiate authentic content from AI media, raising concerns around misinformation, spam, and user trust. In response, TikTok has introduced a series of AI literacy measures aimed at helping users better understand AI generated content while strengthening its efforts to detect and label such media. The initiative combines educational resources, expanded AI content labeling, and improved detection systems. It is all part of a bigger push to be more […]

GPT-6 and GPT-5.7 Leaks Point to an August Launch, 10T Scale & More

By Devanshi Kashyap | July 13, 2026

OpenAI’s upcoming AI models, GPT-5.7 and GPT-6’s leaks have started a discussion across the tech industry. Several reports claim the company is preparing one of its biggest model upgrades yet, with rumours pointing to an August launch along with major improvements in reasoning, memory, context handling, and AI agent capabilities. None of this is official yet, but people are expecting these new models to go way beyond basic chatbots and actually help with research, coding, business automation, and complex problem solving. However, the reported details should be treated as speculation until the company officially confirms them. GPT-6/GPT-5.7 Rumoured Upgrades and […]

Meta Increases Hyperion AI Data Center Investment to $50 Billion as Compute Demand Surges

By Khwaish Manwani | July 13, 2026

Meta has broadened its artificial intelligence infrastructure vision, declaring that its Hyperion data center supercluster in Louisiana has increased in cost. It has risen to $50 billion and will eventually scale to 5 gigawatts. The change is nearly double earlier calculations and depicts the company’s focus on acquiring computing power to support upcoming artificial intelligence models. As competitiveness increases among hyperscalers, the expansion also emphasizes how states are using tax exemption and energy resources to appeal to large-scale AI infrastructure projects. It is also rapidly growing to fulfill Meta’s long-term AI ambition. Why is Meta Expanding the Hyperion Project? Meta’s […]

Microsoft AI CEO Nadella’s “Reverse Information Paradox” Is a Warning for Enterprise AI

By Khwaish Manwani | July 13, 2026

Artificial intelligence consists of effectiveness, quick decision-making, and smart enterprise operations. Microsoft’s CEO, Satya Nadella believes that with all these benefits comes a new issue that organizations cannot feign ignorance about. He states it as the reverse information paradox, which is an idea that businesses may give out their valuable internal knowledge to AI systems to get substantial results. While artificial intelligence becomes more effective when it understands how an enterprise operates, the same process could hand out the enterprise workflows, decision-making patterns, and expertise that give companies their competitive advantage. What is the Reverse Information Paradox? Conventionally, a massive […]

AI & Tech News