Greetings! Welcome to TechScape. I’m your host, Blake Montgomery.
A fierce battle is underway between tech giants and startups to seize control of the next era of artificial intelligence.
Amazon has taken legal action against Perplexity AI, a notable AI startup, regarding a shopping functionality in its browser that enables users to automate orders. Amazon alleges that Perplexity AI has been covertly accessing customer accounts, making the AI’s actions appear as though they were human browsing.
This dispute underscores the emerging discussions about regulating the increasing utilization of AI agents, autonomous digital assistants powered by AI, and their interactions with online platforms. Perplexity is developing a browser named Comet, which incorporates an AI agent, but Amazon is opposed to allowing Comet to facilitate user shopping. This opposition is grounded in factual concerns. Microsoft found that research simulations revealed AI agents are often vulnerable to manipulation while shopping.
This situation raises numerous questions. Are Perplexity’s agents reckless buyers posing significant security threats, or is Amazon attempting to stifle its emerging competitors? Whose interests do these semi-autonomous AI agents serve—those of the customer or the manufacturer’s? And who will bear the responsibility for any misconduct? The future iteration of AI could significantly influence legal outcomes.
Perplexity is not necessarily a champion for the average consumer against Amazon’s overwhelming influence. The startup has secured $1.5 billion at a $20 billion valuation, as reported by TechCrunch. Throughout this process, the company has accumulated textual data with little regard for the rights of content creators, trained various AI systems, and subtly bypassed clear restrictions against unauthorized data scraping. Both Forbes and Wired have highlighted instances of the company allegedly plagiarizing their content through deceptive means, as noted by The Verge. We have put together an extensive list of controversies surrounding Perplexity.
The firm is eager for market share and profits, seemingly willing to trample over any competitor—big or small—to achieve its goals. Jeff Bezos, founder of Amazon, might resonate with this mindset, often criticized for his own ruthlessness. Notably, he has invested in Perplexity on two occasions.
A Future Full of Challenges Emerges
Recently, AI has made significant advancements in music and international relations. My colleague Aisha Down reports:
This week, three songs created by artificial intelligence reached the top of music charts, including Spotify and Billboard.
According to Spotify, Breaking Rust’s “Walk My Walk” and “Livin’ on Borrowed Time” topped the U.S. “Viral 50” list, while Dutch song “We Say No, No, No to an Asylum Center,” an anti-immigrant anthem by JW “Broken Veteran,” dominated the global viral chart. Furthermore, “Breaking Rust” secured a position in the top five worldwide.
In a recent study from streaming platform Deezer, it was revealed that 50,000 AI-generated songs are uploaded daily, accounting for 34% of all music shared on the platform.
Perhaps the next wave will be in podcasting. The AI startup Inception Point produces 3,000 episodes each week. As reported, their distribution network boasts 400,000 subscribers, resulting in 12 million episode downloads, with each episode priced at $1. Apple Music and Spotify together host approximately 175,000 AI-generated podcast episodes.
On the diplomatic front, AI firm Anthropic revealed it had identified and prevented a largely automated cyberattack from Chinese state-sponsored hackers. Aisha again:
US-based Anthropic announced that its coding tool, Claude Code, was utilized by a Chinese state-supported group to target 30 organizations globally in September, leading to “several successful breaches.”
After newsletter promotion
According to a blog post by Anthropic, this incident represents a “significant escalation” compared to prior AI-driven attacks they have monitored, as Claude operated mostly autonomously. During the attack, 80-90% of operations were conducted without human intervention.
“This attacker has reportedly conducted what we believe is the first documented large-scale cyberattack performed entirely without human input,” the report states.
The emergence of automated threats is becoming increasingly prevalent. Even if one cyberattack is thwarted, more could arise in rapid succession. If a single AI-generated album is removed from a platform, several others could fill the void immediately. In the not-so-distant future, we might find ourselves navigating through constant tumultuous changes.
Discover more about AI and politics
Explore more about AI and the environment
Learn more about AI bubbles
Source: www.theguardian.com
