Cybersecurity Looks to AI

Shanny Basar

The spread of ransomware holding computers to ransom last week highlights the impact of weaknesses in cybersecurity as British hospitals were forced to cancel operations. Corvil, which provides data analytics for electronic trading, is launching “Cara” which uses machine learning algorithms to safeguard electronic trading businesses without affecting performance.

Corvil said Cara is the first virtual security expert to address cyber security in financial markets.

The G7 Finance Ministers said in a statement at the weekend that the increasing digitalization of financial services has led to the need for effective approaches for the assessment of cybersecurity at the firm and sector level

“Given the rapidly evolving nature of cyber risks, effective cyber security assessment requires that current assessment approaches should be enhanced and be complemented by practices that are tailored to bolster cyber resilience, including regular cyber exercises and simulations as well as consideration of how to most effectively leverage penetration tests,” said the G7.

As a result the G7 Cyber Expert Group has been asked to develop a set of high level fundamental elements for effective assessment of cyber security by October 2017. The G7 continued that it is following the development of a cyber insurance market and the ongoing work by The Organisation for Economic Co-operation and Development n this area.

“No point of cyberspace can be absolutely secure as long as cyber threats persist in the surrounding environment; our drive to strengthen the financial system against cyber attacks can achieve maximum results only if accompanied by measures that reduce the level of insecurity in cyberspace as a whole,” said the G7.

David Murray, chief business development officer at Corvil, told Markets Media that many firms think of trading infrastructure as secure isolated environments but they are vulnerable to malware in data feeds, connections to external venues and compromised user accounts.

He said: “81% of cyber attacks are a result of compromised user accounts. Can you spot a user logging in at an unusual location ?”

Corvil “Cara” uses machine learning algorithms to determine and benchmark the normal behavior of activity seen on the network to identify changes in normal patterns and which need further investigation. A daily cyber security intelligence report includes a cyber risk assessment score so management and security teams can prioritise which anomalies require most urgent action, as they are freed up from the need to manually analyse large amounts of data. Firms can also set their own circuit breakers so trigger a halt in trading on the occurrence of certain events.

Murray added that machine learning allows Cara to makes sense of the specific user and business context of all trading sessions, market data streams and other traffic and reduce false positives as the algorithms near more about a firm’s patterns of behaviour.

Dan Cummins, senior analyst, security at 451 Research, said in a statement: “Corvil Virtual Security Expert gives security teams a relatively quick way to extend automated risk assessments, which combine machine-learning anomaly detection and threat detection analytics, to electronic trading networks.”

Corvil Cara will be generally available June 30 and will not require additional infrastructure for existing Corvil analytics customers.

The need for solutions for cybersecurity and market surveillance was highlighted this month when Cinnober Financial Technology, which supplies financial technology to marketplaces and clearinghouses, said it was acquiring Ancoa Software. The UK-based firm provides contextual surveillance and analytics for exchanges, regulators, buyside and sellside firms.

Veronica Augustsson, chief executive of Cinnober, said in a stament: “By bringing Ancoa into Cinnober we are now adding new competences as well as deepening our leading product portfolio targeting banks and brokers as well as exchanges and clearinghouses”, says We’re also adding approximately ten additional customers to our client base, within a highly interesting segment of our business niche driven by regulatory demands.”

Related articles

  1. Algos, Post-Trade Top FCM Concerns

    Instinet's Anushree Laturkar says investing in research and analysis is key to improving client outcomes.

  2. AI can optimize algorithms to better recognize and adapt to real-time market conditions.

  3. WEX Options Algo Built for Market Complexity

    Due diligence process has not kept pace with the proliferation of algorithmic trading.

  4. By Greta Zhou and Andy Cheung, APAC AES, Credit Suisse

  5. Broker's new platform generates execution strategies that leverage client-specific trading data.