How to protect money, clients and reputation of the company using sensing technologies

Only for the first half of the year 2017 there were 925 large-scale leaks of confidential information that is 10% more, than for the same period in 2016 – such data are provided by experts of Infowatch company. And the majority of incidents (about 56%) including connected with leak of images took place because of the staff of the companies. For example, in July as a result of an internal error the Australian travel agency Flight Centre transferred passport data of hundreds of clients to strangers. And the Swedish transport agency at all accidentally sent to e-mail mailing with base of all vehicles in the country, including names and photos of owners of cars - even citizens who are in the protection program of witnesses.

Leaks in the form of the scanned copies, photos and screenshots quite often happen to confidential information also in Russia. So, according to Solar Dozor, 31% of such incidents in the Russian companies are the share of the financial sector, 21% of cases occur in the fuel and energy sector, 17% of leaks are the share of state structures and 11% - on retail. About 8% of leaks happen in the industries of mechanical engineering and a MIC, but exactly there specialists note high dynamics, - the number of similar incidents constantly grows in these spheres.

Partially competitive intelligence is the reason of such leaks. Lists of clients, the internal management reporting and operational plans flow away. Other reason - internal fraud and use of information for personal reasons. It is possible to assess such damages only approximately – and these are about 3% of revenue of the companies annually, – Vasily Lukinykh, the business development manager of Solar Dozor of Solar Security company notes.

According to a research of SearchInform company of 31% of the Russian specialists in information security from the large Russian companies noted sharpness of a problem of date leak in the form of images.

Under the threat there is more than a third of data of business: analysts of SearchInform considered that the volume of the scanned copies averages about 30% of all documents. In a public sector it is about 41.5% of documents, in retail — 17%, in the field of services — 23%, and in banks and the telecom sphere approach 45%.

If the malefactor has data access, then only the DLP system checking all outgoing communications regarding transfer of data, critical for business, including on images can protect information, - the leading analyst of SearchInform Alexey Parfentyev says.

Sensing technologies of ABBYY as a part of DLP solutions

The leading Russian developers DLP claim that full protection of crucial assets requires use of OCR technologies. They help business to prevent leak of important documents in the form of scans and photos in time – passports, car driver licenses and even drawings.

For operation of the OCR module in DLP several factors have critical value. First of all, it is performance, accuracy of recognition, a possibility of document classification and support of different languages. According to specialists on information security, technologies of ABBYY have advantages in these directions. They are used as the system of recognition by the largest Russian DLP manufacturing: InfoWatch, Solar Security, SearchInform and Zecurion which general share exceeds 75% of the market DLP (according to analysts of Anti-Malware company). Besides, technologies ABBYY use the companies in the solutions Falcongaze and Symantec.

During the work with texts in English of technology of ABBYY quite strongly win on speed. With Russian there is a noticeable overweight on quality of recognition, - Alexander Akimov, the CEO of FalconGaze comments.

As a result of own tests specialists of SearchInform were convinced that the OCR module constructed based on technologies of ABBYY for 10-12% will recognize the literal text more precisely and 30% better processes difficult images. At the same time algorithms of ABBYY use possibilities of modern processors, and some tasks are carried out 3-4 times quicker than standard OCR based on open technologies. Also the technology of ABBYY allows to make document classification on appearance, contents and other signs: select in the general flow of documents of the passport, credit cards, car driver licenses and so on.

According to InfoWatch company, 90% of customers of the DLP solution InfoWatch Traffic Monitor apply sensing technologies of ABBYY quickly to find the confidential information transferred in a type of images.

The most part of problems in DLP is solved by the analysis of the text: for example, on a key word, taking into account synonyms and different forms, on phrases, attributes or structure of documents. "If the text in the document is recognized incorrectly, a system can just pass and not execute it the main function, - Andrey Isaev, the director of products of ABBYY group says. - And, of course, good support of Cyrillics as the majority of the confidential data which are subject to control using DLP systems contain in documents in Russian is important for the Russian customers.

Investigations of incidents. As the companies prevented leak using OCR ABBYY

The technology of optical text recognition of ABBYY is extremely demanded in the companies where graphic files, in particular, in design and design offices, on industrial enterprises are sensitive to information leak. Also the module of ABBYY is often used where work with the scanned documents is conducted: for example, in insurance companies and banks. Together with ABBYY we can prevent leak not only text or graphic documents, but also the paper digitized documents, - Vladimir Ulyanov, the head of analytical center Zecurion Analytics noted.

How do customers of DLP solutions use the OCR module in practice? Experts of cybersecurity gave a number of examples. So, in one Russian company the security system from the organizational point of view was not bad built: all "excess" communication channels are closed, personal devices are prohibited, exits in social networks and so on are blocked. The data access was strictly regulated. Despite precautions of a security service, one sales manager, having no other options any, methodically did screenshots from CRM and thus printed base of clients. Thanks to use of the OCR module based on technologies of ABBYY, at printing of the first screenshot the security service received the notification. On the fiftieth page a warning was received by the head of the safety department. Of course, left the manager, and leak was prevented.

Other example belongs to construction company. With participation in the tender of one of architects main competitors bribed. They promised it the large amount that it "will merge" the project for the tender. The unreliable employee agreed, but when he tried to send sketches of the project by e-mail, the OCR module based on technologies of ABBYY transformed the graphic file and selected in the text relating to confidential documents is mute. Further the Zecurion Traffic Control system (Zgate) analyzed it and blocked file transfer.

Quite often theft of information belongs to personal data. Using DLP "SearchInform" with the OCR module based on technologies of ABBYY, the security service specialist of hotel found out that someone sends scans of passports of visitors of external mail. Investigation showed that the employee a reception sold them on the specialized websites. So it was succeeded to stop at an early stage activity of the swindler in a workplace.

Of course, if to speak about leaks, it is necessary to mention the financial sector. In one bank the security service detected in draft copies of personal mail of the employee of the photo of credit cards of clients. It became possible thanks to the fact that the DLP system of SearchInform "tightened" a photo, and technologies of ABBYY recognized and analyzed the text when the specialist came on mail from the working PC. Investigation showed that the employee thus merged data to swindlers, and for "delivery" of the victims the percent relied the employee of the bank.

However the most unexpected case of leak was nearly organized by quite loyal employee. The girl decided to brag of unusual manicure in social networks and made a photo of "marigold" against the background of the documents which are a trade secret of the company. The Solar Dozor system intercepted the photo, thanks to the fact that the OCR module recognized the document on a background, and data transmission was blocked.

It is prepared with assistance of ABBYY company

236

Notes