namespace Google\Site_Kit_Dependencies\GuzzleHttp\Promise; /** * Get the global task queue used for promise resolution. * * This task queue MUST be run in an event loop in order for promises to be * settled asynchronously. It will be automatically run when synchronously * waiting on a promise. * * * while ($eventLoop->isRunning()) { * GuzzleHttp\Promise\queue()->run(); * } * * * @param TaskQueueInterface $assign Optionally specify a new queue instance. * * @return TaskQueueInterface * * @deprecated queue will be removed in guzzlehttp/promises:2.0. Use Utils::queue instead. */ function queue(\Google\Site_Kit_Dependencies\GuzzleHttp\Promise\TaskQueueInterface $assign = null) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::queue($assign); } /** * Adds a function to run in the task queue when it is next `run()` and returns * a promise that is fulfilled or rejected with the result. * * @param callable $task Task function to run. * * @return PromiseInterface * * @deprecated task will be removed in guzzlehttp/promises:2.0. Use Utils::task instead. */ function task(callable $task) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::task($task); } /** * Creates a promise for a value if the value is not a promise. * * @param mixed $value Promise or value. * * @return PromiseInterface * * @deprecated promise_for will be removed in guzzlehttp/promises:2.0. Use Create::promiseFor instead. */ function promise_for($value) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Create::promiseFor($value); } /** * Creates a rejected promise for a reason if the reason is not a promise. If * the provided reason is a promise, then it is returned as-is. * * @param mixed $reason Promise or reason. * * @return PromiseInterface * * @deprecated rejection_for will be removed in guzzlehttp/promises:2.0. Use Create::rejectionFor instead. */ function rejection_for($reason) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Create::rejectionFor($reason); } /** * Create an exception for a rejected promise value. * * @param mixed $reason * * @return \Exception|\Throwable * * @deprecated exception_for will be removed in guzzlehttp/promises:2.0. Use Create::exceptionFor instead. */ function exception_for($reason) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Create::exceptionFor($reason); } /** * Returns an iterator for the given value. * * @param mixed $value * * @return \Iterator * * @deprecated iter_for will be removed in guzzlehttp/promises:2.0. Use Create::iterFor instead. */ function iter_for($value) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Create::iterFor($value); } /** * Synchronously waits on a promise to resolve and returns an inspection state * array. * * Returns a state associative array containing a "state" key mapping to a * valid promise state. If the state of the promise is "fulfilled", the array * will contain a "value" key mapping to the fulfilled value of the promise. If * the promise is rejected, the array will contain a "reason" key mapping to * the rejection reason of the promise. * * @param PromiseInterface $promise Promise or value. * * @return array * * @deprecated inspect will be removed in guzzlehttp/promises:2.0. Use Utils::inspect instead. */ function inspect(\Google\Site_Kit_Dependencies\GuzzleHttp\Promise\PromiseInterface $promise) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::inspect($promise); } /** * Waits on all of the provided promises, but does not unwrap rejected promises * as thrown exception. * * Returns an array of inspection state arrays. * * @see inspect for the inspection state array format. * * @param PromiseInterface[] $promises Traversable of promises to wait upon. * * @return array * * @deprecated inspect will be removed in guzzlehttp/promises:2.0. Use Utils::inspectAll instead. */ function inspect_all($promises) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::inspectAll($promises); } /** * Waits on all of the provided promises and returns the fulfilled values. * * Returns an array that contains the value of each promise (in the same order * the promises were provided). An exception is thrown if any of the promises * are rejected. * * @param iterable $promises Iterable of PromiseInterface objects to wait on. * * @return array * * @throws \Exception on error * @throws \Throwable on error in PHP >=7 * * @deprecated unwrap will be removed in guzzlehttp/promises:2.0. Use Utils::unwrap instead. */ function unwrap($promises) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::unwrap($promises); } /** * Given an array of promises, return a promise that is fulfilled when all the * items in the array are fulfilled. * * The promise's fulfillment value is an array with fulfillment values at * respective positions to the original array. If any promise in the array * rejects, the returned promise is rejected with the rejection reason. * * @param mixed $promises Promises or values. * @param bool $recursive If true, resolves new promises that might have been added to the stack during its own resolution. * * @return PromiseInterface * * @deprecated all will be removed in guzzlehttp/promises:2.0. Use Utils::all instead. */ function all($promises, $recursive = \false) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::all($promises, $recursive); } /** * Initiate a competitive race between multiple promises or values (values will * become immediately fulfilled promises). * * When count amount of promises have been fulfilled, the returned promise is * fulfilled with an array that contains the fulfillment values of the winners * in order of resolution. * * This promise is rejected with a {@see AggregateException} if the number of * fulfilled promises is less than the desired $count. * * @param int $count Total number of promises. * @param mixed $promises Promises or values. * * @return PromiseInterface * * @deprecated some will be removed in guzzlehttp/promises:2.0. Use Utils::some instead. */ function some($count, $promises) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::some($count, $promises); } /** * Like some(), with 1 as count. However, if the promise fulfills, the * fulfillment value is not an array of 1 but the value directly. * * @param mixed $promises Promises or values. * * @return PromiseInterface * * @deprecated any will be removed in guzzlehttp/promises:2.0. Use Utils::any instead. */ function any($promises) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::any($promises); } /** * Returns a promise that is fulfilled when all of the provided promises have * been fulfilled or rejected. * * The returned promise is fulfilled with an array of inspection state arrays. * * @see inspect for the inspection state array format. * * @param mixed $promises Promises or values. * * @return PromiseInterface * * @deprecated settle will be removed in guzzlehttp/promises:2.0. Use Utils::settle instead. */ function settle($promises) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Utils::settle($promises); } /** * Given an iterator that yields promises or values, returns a promise that is * fulfilled with a null value when the iterator has been consumed or the * aggregate promise has been fulfilled or rejected. * * $onFulfilled is a function that accepts the fulfilled value, iterator index, * and the aggregate promise. The callback can invoke any necessary side * effects and choose to resolve or reject the aggregate if needed. * * $onRejected is a function that accepts the rejection reason, iterator index, * and the aggregate promise. The callback can invoke any necessary side * effects and choose to resolve or reject the aggregate if needed. * * @param mixed $iterable Iterator or array to iterate over. * @param callable $onFulfilled * @param callable $onRejected * * @return PromiseInterface * * @deprecated each will be removed in guzzlehttp/promises:2.0. Use Each::of instead. */ function each($iterable, callable $onFulfilled = null, callable $onRejected = null) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Each::of($iterable, $onFulfilled, $onRejected); } /** * Like each, but only allows a certain number of outstanding promises at any * given time. * * $concurrency may be an integer or a function that accepts the number of * pending promises and returns a numeric concurrency limit value to allow for * dynamic a concurrency size. * * @param mixed $iterable * @param int|callable $concurrency * @param callable $onFulfilled * @param callable $onRejected * * @return PromiseInterface * * @deprecated each_limit will be removed in guzzlehttp/promises:2.0. Use Each::ofLimit instead. */ function each_limit($iterable, $concurrency, callable $onFulfilled = null, callable $onRejected = null) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Each::ofLimit($iterable, $concurrency, $onFulfilled, $onRejected); } /** * Like each_limit, but ensures that no promise in the given $iterable argument * is rejected. If any promise is rejected, then the aggregate promise is * rejected with the encountered rejection. * * @param mixed $iterable * @param int|callable $concurrency * @param callable $onFulfilled * * @return PromiseInterface * * @deprecated each_limit_all will be removed in guzzlehttp/promises:2.0. Use Each::ofLimitAll instead. */ function each_limit_all($iterable, $concurrency, callable $onFulfilled = null) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Each::ofLimitAll($iterable, $concurrency, $onFulfilled); } /** * Returns true if a promise is fulfilled. * * @return bool * * @deprecated is_fulfilled will be removed in guzzlehttp/promises:2.0. Use Is::fulfilled instead. */ function is_fulfilled(\Google\Site_Kit_Dependencies\GuzzleHttp\Promise\PromiseInterface $promise) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Is::fulfilled($promise); } /** * Returns true if a promise is rejected. * * @return bool * * @deprecated is_rejected will be removed in guzzlehttp/promises:2.0. Use Is::rejected instead. */ function is_rejected(\Google\Site_Kit_Dependencies\GuzzleHttp\Promise\PromiseInterface $promise) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Is::rejected($promise); } /** * Returns true if a promise is fulfilled or rejected. * * @return bool * * @deprecated is_settled will be removed in guzzlehttp/promises:2.0. Use Is::settled instead. */ function is_settled(\Google\Site_Kit_Dependencies\GuzzleHttp\Promise\PromiseInterface $promise) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Is::settled($promise); } /** * Create a new coroutine. * * @see Coroutine * * @return PromiseInterface * * @deprecated coroutine will be removed in guzzlehttp/promises:2.0. Use Coroutine::of instead. */ function coroutine(callable $generatorFn) { return \Google\Site_Kit_Dependencies\GuzzleHttp\Promise\Coroutine::of($generatorFn); } Artem Lyashanov: How artificial intelligence changes the rules of the game in banking compliance – Guitar Shred

Guitar Shred

Artem Lyashanov: How artificial intelligence changes the rules of the game in banking compliance

Escrito por

wordpress_administrator

em

Can a modern bank survive in the conditions of regulatory pressure, endless lists of PBC and threats of financial criminals, relying only on manual labor?

In 2026, the answer is obvious: no. While the industry faces the outflow of young talent and the complexity of legislation, artificial intelligence (AI) is becoming the lifeline that not only automates reports, but also fundamentally changes approaches to security and transparency.

Artem Lyashanov, an expert in the field of fintech innovations, emphasizes that the transition to AI compliance is a necessity for maintaining competitiveness. In this article, we will consider how algorithms help in fraud detection, document management and risk forecasting, and we will also understand real cases of world banks.

Detection of fraud and monitoring of transactions

Outdated methods often create a huge amount of noise, due to which real threats remain unnoticed among thousands of false positives. This not only overloads security departments, but also annoys honest customers whose accounts are blocked unnecessarily.

AI offers a fundamentally different approach, for example, instead of static lists, it uses machine learning to analyze huge arrays of data in real time. This allows not only to catch fraudsters faster, but also to significantly reduce the number of errors, making life easier for honest users.

Banks get the opportunity to see anomalies where the human eye or old software cannot see anything.

Key advantages of implementation:

  • Algorithms are constantly being improved, learning new tactics of cybercriminals without the need to manually update the rules;
  • AI more accurately distinguishes atypical customer behavior from real theft of funds;
  • Transaction processing and issuing a verdict on its security take place in milliseconds.

HSBC uses AI to monitor credit card transactions, which allows it to block suspicious activity before it causes significant damage. Standard Chartered has implemented an anti-money laundering (AML) system that analyzes communications between senders and receivers around the world.

Automation of PBC documentation and lists

One of the most difficult stages of the audit is the preparation of PBC (Prepared By Client) lists – documents that the client must provide for verification.

Previously, this process required hundreds of hours of searching for papers, correspondence and endless reconciliations, which demotivated employees and lengthened the period of inspections. The human factor with such a volume of work inevitably led to errors.

Instead of flipping through hundreds of pages, a specialist can simply request the necessary figures or contract terms from the system. This allows the team to focus on validating data and making strategic decisions.

New generation instruments:

  • DocuMine allows you to upload arrays of documents and instantly receive answers to compliance questions with links to specific paragraphs;
  • UpLink automates PBC requests, centralizing the process of collection and preliminary verification of documents;
  • Intelligent reporting – automatic generation of text reports based on analyzed data.

After the implementation of such systems, banks note a radical reduction in preparation time for inspections. Freed resources are directed to a deep analysis of risks, which significantly increases the quality of the audit.

Risk forecasting

Thanks to predictive analytics, banks no longer just record the facts of violations, but try to predict their probability on the basis of indirect signs.

Analyzing market reports and even social networks, AI helps to form a clear picture of potential threats and future regulatory changes.

In addition, routine employee requests are now processed by intelligent chatbots, which significantly speeds up internal processes. They help to quickly understand complex procedures, fill out forms correctly, or get clarifications on new laws such as GDPR.

Development directions:

  • Virtual assistants advising staff on regulatory issues 24/7;
  • Modeling scenarios of legislative changes and their influence on the bank’s business processes;
  • Constant monitoring of systems for compliance with security standards without waiting for scheduled inspections.

“Banks using predictive models are the first to adapt to the new rules of the game, avoiding huge fines and reputational losses. AI becomes the foundation of trust between the financial institution, the regulator and the client.”

Artem Lyashanov

Comentários

Deixe um comentário

O seu endereço de e-mail não será publicado. Campos obrigatórios são marcados com *

Mais posts