Shares in numerous software companies have dropped after artificial intelligence firm Anthropic released new agentic AI tools, leaving some investors worried about the future of traditional software-as-a-service (SaaS) firms.
The S&P 500 software and services index lost around 8 per cent of its value last week, and has shed around $US1 trillion ($1.4 trillion) since 28 January in a selloff which some have even dubbed “software-mageddon” or the “SaaSpocalypse”.
Selloffs impacted companies including Atlassian, Salesforce, ServiceNow, Microsoft, SAP, and Workday, which have increasingly faced concerns about the impact accessible and powerful AI agents may have on demand for their software.
The share market dips came after Anthropic unveiled new plugins for its agentic Claude Cowork platform which allowed users to tailor the tool for use in particular sectors such as finance, legal, marketing, and data analysis.
Technology analyst Lian Jye Su from Omdia said that because Claude Cowork operated autonomously and could be interacted with using natural language, it was “a lot more advanced than traditional SaaS solutions that are business process oriented”.
“Claude Cowork is also seen to be the perfect alternative to build custom tools that can break down data silos and automate complex workflows,” he told Information Age.
“Investors see Cowork as a direct threat to incumbents like Salesforce, Workday, or ServiceNow, whose models rely on ongoing human interaction with their platforms.”
Software stocks regained some of their losses on Friday and Monday, when both the S&P 500 and Nasdaq Composite rallied — but this was partly due to chip manufacturers seeing boosts from positive sentiment around AI spending.
The S&P 500 software and services index remains down around 25 per cent since its peak in October 2025.
Su said he believed software stocks would see “a long-term correction for feature-focused or easily replicable SaaS”.
“AI advancement could trigger existential problems for undifferentiated and overvalued SaaS offerings,” he said.
“However, high-quality platforms with deep integrations may still benefit from the AI advancement.
“The winners will be those who evolve into AI-native ecosystems, but many legacy players risk irrelevance if they can't.”
‘Others think software is dead’, says Atlassian CEO
Australian software giant Atlassian — whose share price on the Nasdaq has dropped around 70 per cent over the past 12 months — still saw AI as a boon for its products and customers, co-founder and CEO Mike Cannon-Brookes told investors on Friday.
“I’m convinced AI is great for Atlassian. Others think software is dead,” he wrote in a letter to shareholders.
Speaking on an earnings call, the executive later admitted the Atlassian team was “frustrated” by the company’s share price slump, but said he was “incredibly bullish” on AI.
Atlassian CEO Mike Cannon-Brookes says the software firm doesn't see AI as 'the challenge that others do'. Image: Atlassian / Supplied
Asked by analysts about the impact of AI tools such as Anthropic’s Cowork, which was first released in January, Cannon-Brookes said Atlassian used Anthropic’s AI coding tools in its own engineering teams and would continue to integrate AI services into its own applications.
“There is significant value, I think, between our offerings and those offerings,” he said.
“So we don’t see that as being perhaps the challenge that others do out there.
“There’s a great partnership opportunity there, and we continue to explore that, we continue to use their offerings really strongly internally.”
Cannon-Brookes argued Atlassian’s apps provided agentic tools with data, context, and places for information exchange, which allowed for better human-AI interaction within business workflows.
Atlassian, which has its own AI tool called Rovo, recently acquired an AI browser company and an AI-focused developer productivity platform to bolster its AI offerings.
Anthropic and OpenAI release more agentic tools
After spooking software shares with its Claude Cowork plugins, Anthropic also released its latest flagship model Claude Opus 4.6 on Friday.
The new model has greater coding and reasoning skills, and sustained agentic tasks for longer, the company said.
Claude Opus 4.6 could also use these improved abilities within Cowork to complete tasks such as “running financial analyses, doing research, and using and creating documents, spreadsheets, and presentations”, Anthropic said.
Also on Friday, ChatGPT maker OpenAI launched a new platform for building, deploying, and managing AI agents, dubbed Frontier.
Matt Comyn, CEO of Australia’s Commonwealth Bank — which has a strategic partnership with OpenAI — said in a statement that the financial giant had “early exposure to elements of the Frontier platform” and could see “how it can support the building, deployment and management of AI".
OpenAI is reportedly hiring hundreds of technical consultants known as forward-deployed engineers, to help enterprise customers integrate AI agents as it faces increased competition from Anthropic.
Anthropic was founded in 2021 by seven former employees of OpenAI.
Despite an increasing number of Australian businesses adopting AI tools, a recent survey found only one in seven companies had seen revenue gains as a result.
