Top Story: Alaan Raises $48 Million to Grow Its AI Finance Tool in the Middle East
Alaan is a company that helps businesses in the Middle East manage their spending with the help of AI. It was founded by two former McKinsey consultants, Parthi Duraisamy and Karun Kurien, after they experienced firsthand the difficulties of managing company expenses in the region. The company just raised $48 million from big investors like Peak XV (formerly Sequoia India), Y Combinator, and others. Alaan’s tool saves finance teams time by using AI to match receipts, track expenses, and handle tax paperwork. Big companies like Careem, Tabby, and Lulu Group already use it. Even though it took time to get government approval in places like the UAE and Saudi Arabia, Alaan is now growing fast. The company says it is already profitable, earning $10 million in revenue after spending only $5 million. With the new funding, Alaan plans to hire more people, improve its AI features, and grow its business in Saudi Arabia. Investors believe in the company because it is efficient and has strong customer demand. Google’s AI Finds 20 Security Bugs in Popular Software
Google’s new AI tool, called Big Sleep, has found 20 security problems in well-known free software like FFmpeg and ImageMagick. The tool was built by DeepMind and Google’s top hacker team, Project Zero. While people still need to double-check the findings before reporting them, the AI was able to spot and test the bugs on its own. Google isn’t sharing details yet because the issues haven’t been fixed. This shows AI can really help find hidden problems in software. But there are still worries. Some software makers say AI tools sometimes report bugs that don’t exist. Still, Big Sleep is getting attention because smart people are behind it, and the tool has proven it can work well. Other similar tools, like RunSybil, are also trying to make bug hunting faster and easier using AI.
OpenAI has updated ChatGPT to stop giving direct advice on serious personal matters like breakups. Instead of telling users what to do, ChatGPT will now guide them through questions and options to help them make their own decisions. This change was made after concerns that the AI was acting too agreeably or giving advice in emotional situations without enough care. To improve safety, OpenAI is also adding gentle reminders for users to take breaks during long chats. The company has formed a new advisory group of mental health experts, doctors, and researchers to help ensure the tool behaves responsibly. More than 90 doctors from over 30 countries reviewed how ChatGPT responds in sensitive situations to make sure it helps users without crossing boundaries. OpenX, a smaller advertising technology company, is suing Google in a U.S. court. OpenX says Google used unfair tactics to take control of online ad sales. It claims Google forced websites to use its tools and rigged ad auctions so its products would win. Because of this, OpenX had to shut down part of its business in 2019. Now, the company wants Google to pay damages and stop these actions. This lawsuit follows a recent court win by the U.S. government, which found Google had broken competition laws in the ad market. OpenX also accuses Google of using secret programs and giving itself unfair advantages. If OpenX wins, it could lead to big changes in how Google runs its ad business or even force it to break up part of it. Perplexity AI Accused of Bypassing Website Blocks to Scrape Content Cloudflare has accused AI startup Perplexity of secretly scraping content from websites that explicitly said they did not want to be scraped. According to Cloudflare, Perplexity ignored standard internet rules set by sites using robots.txt files and used tricks like changing its bot identity and disguising itself as Google Chrome to get around these restrictions. Cloudflare says this behaviour was seen across thousands of websites and millions of daily requests, and it confirmed the activity after complaints from customers. The company has since blocked Perplexity’s bots from its verified list and introduced new measures to stop further scraping. Perplexity denies the claims, calling Cloudflare’s blog post a “sales pitch” and saying the screenshots shared show no content was accessed. However, this is not the first time the startup has been accused of scraping content without permission. Last year, Wired and other media outlets said Perplexity was plagiarising their work, and during a public interview, the company’s CEO struggled to define plagiarism. Meanwhile, Cloudflare has taken a stronger stance on AI scrapers in recent months, even launching tools to help publishers charge AI firms or block their bots entirely. AI Tool of the DayDeepBrain
What it does:
AI Word of the DayModel HallucinationDefinition: Why it matters: That’s it for today, guys. AI moves fast, but we’ve got you faster. Catch you next drop Staff Writer, The Algorithm Daily |
ChatGPT Shouldn’t End Relationships
OpenAI has updated ChatGPT to stop giving direct advice on serious personal matters like breakups.
