Skip to content

AvalonX Gets Up To $10M To Sell Small Firms AI Hacking Tests

Ridown Capital, run by Onecom’s founders, is backing US startup AvalonX, whose AI agents run simulated cyberattacks for small firms from $249 a month.

AvalonX Gets Up To $10M To Sell Small Firms AI Hacking Tests
Image courtesy: AvalonX

AvalonX, a cybersecurity startup from Alpharetta, Georgia, has emerged after 18 months of working in stealth, out of public view, with up to $10M from Ridown Capital. The company uses AI to attack its customers’ systems, with their permission, the way a criminal would, so small and midsize businesses can find and fix weak spots. AvalonX has not disclosed the type of round or its valuation.

Ridown Capital is the investment arm of Ridown Group. The group is led by Aaron Brown and Darren Ridge, who co-founded Onecom, one of Britain’s largest independent business communications providers. AvalonX says Onecom serves more than 65,000 small and midsize businesses, and that Ridown Group’s telecoms, cybersecurity and e-commerce companies bring in more than $500M in revenue in Britain.

Brown said smaller companies are “targeted like enterprises but resourced like startups”, adding: “AvalonX is built for them from the ground up. That’s why we backed it.”

Ridge said he and Brown had spent their careers giving smaller businesses enterprise-level technology at a price that makes sense.

“We’re proud to help bring it to businesses that have historically been priced out of this level of testing,” he said.

Hacking Tests On Subscription

Large companies can afford to run penetration tests constantly, AvalonX argues, while smaller firms have tended to rely on occasional ones. In a penetration test, security specialists carry out an authorised, simulated attack to find the holes a real criminal could use.

AvalonX’s Artemis platform turns that into continuous testing, sold as a monthly subscription. According to the company, a team of AI agents, software programs that act on their own, copies an attacker’s steps, from scouting a company’s systems to breaking in and moving deeper inside its network, and each agent shares what it finds with the others. Human penetration testers then check the findings that need expert judgement, so customers receive a prioritised list of confirmed problems instead of a flood of automated alerts.

The agents run on Aether, AvalonX’s own security AI model. The company says Aether has 400 billion parameters, the internal values a model learns during training, and was trained on nine years of real-world attack data, backed by 60 billion dark-web records.

Customers enter their web domain and choose which systems can be tested, and testing from outside their network needs no software installed. AvalonX says results usually arrive the same day, in plain-language reports with suggested next steps. Plans start at $249 a month, and the company offers a free first test.

Findings can also be mapped to common security standards: SOC 2, an American audit framework for how service companies protect customer data; ISO/IEC 27001, the international standard for information security; the Payment Card Industry Data Security Standard (PCI DSS) for card payments; and the US health data rules known as HIPAA. AvalonX says the mapping gives customers evidence for audits and security reviews.

Small Firms And Ransomware

Ransomware, which locks up a company’s data until a ransom is paid, turns up far more often in breaches at smaller businesses. Verizon’s 2025 Data Breach Investigations Report found ransomware in 88% of breaches at small and midsize businesses, against 39% at larger organisations.

Craig Goodwin, AvalonX’s co-founder and chief executive, was chief information security officer (CISO), the executive in charge of a company’s security, at Fujitsu, Monster and CDK Global, according to the company. He also co-founded Cyvatar, a startup that sells security as a subscription. Rick Valdes and Ben O’Nion, whom AvalonX describes as cybersecurity veterans, join him in leading the business, and the company says the team has more than 50 years of combined experience.

“Large enterprises test constantly. They have red teams, consultants, and budgets to match,” Goodwin said. Red teams are groups paid to attack their own organisation’s defences. Smaller organisations, by his account, test once a year if they can afford it, and “attackers don’t wait eleven months to strike”. He pointed to the likes of a 50-person accounting firm or a regional manufacturer: “We built AvalonX so those businesses get the same quality of testing I relied on as an enterprise CISO.”

AvalonX leaves stealth with backers who have spent their careers selling technology to smaller businesses, and with a monthly subscription aimed at firms that have tended to test once a year, if at all. The approach depends on AI agents handling the attack work, with human testers checking the findings that need judgement.

Add Morning Tick on Google