Anthropic’s expected IPO prospectus warns that its AI models could pose “catastrophic or existential risks to humanity.” But investors also face more immediate uncertainties: whether AI will undermine the intellectual property protections underpinning commercial software, and whether Anthropic could itself be harmed by the technology it helped advance.
The disruption has already affected software companies. Since early 2025, tools including OpenAI’s Operator and Anthropic’s Claude Code have let people create programs using natural-language instructions, a practice known as vibe coding. Anthropic later introduced specialist add-ons for its Claude Cowork agent aimed at legal, regulatory and commercial tasks. These developments contributed to a sharp decline in software-as-a-service shares dubbed the “SaaSpocalypse.”
More recent AI capabilities raise a further concern: models may be able to decompile, reverse-engineer and copy existing programs. Microsoft Xbox chief executive Asha Sharma told staff that games had recently been taken apart and quickly rebuilt to run on devices they were not designed for. She also said major games had been remixed into other titles’ source code.
If AI makes proprietary software easy to reproduce, companies could struggle to charge for products that become effectively non-excludable. That would threaten business models built around subscriptions and individual software sales. AI could also expose vulnerabilities in existing code, although the same tools might help identify and repair security flaws, potentially at significant cost and amid a difficult transition.
The music industry’s experience after the arrival of MP3 technology and Napster provides a comparison. Digital music became cheap to reproduce and easy to share online, prompting widespread piracy. Record companies sued Napster, which went bankrupt in 2001, but legal action did not prevent a profound restructuring of the industry. Established companies including EMI and PolyGram disappeared, while platforms such as Spotify and YouTube emerged.
That history offers no clear guide to which software businesses will ultimately succeed. Napster’s growth-first approach failed, while Apple’s early effort to sell digital music in a way that resembled physical album ownership eventually gave way to subscription and advertising-supported streaming. It took about a decade for streaming to establish itself as a sustainable model, suggesting investors should be cautious about betting on one strategy becoming dominant.
Anthropic also faces copyright litigation that could determine the limits of using AI to reproduce closed-source software. The article argues that the company’s exposure could be much greater than Napster’s if courts rule that cloning proprietary programs infringes intellectual property rights.
There is another difference: AI models can be turned against the companies that develop them. Chinese firms are already reverse-engineering US models and releasing open-source versions, challenging the idea that a closed, exclusive AI frontier can remain a lasting source of value. For investors considering Anthropic, the central question is whether the company can avoid becoming a casualty of the disruption it is helping drive.
Sources: El País Economía.